เกี่ยวกับเรา

คำถามที่พบบ่อย

เรียนรู้คำถามทั่วไปเกี่ยวกับความปลอดภัย การกำกับดูแล และประสิทธิภาพการทำงานด้วย AI ได้อย่างรวดเร็ว

AnyInsight.ai ช่วยให้องค์กรนำแอปพลิเคชัน AI ไปใช้และขยายการใช้งานได้อย่างมีประสิทธิภาพ ภายในสภาพแวดล้อมที่ปลอดภัยและควบคุมได้

No. AnyInsight.ai features de-identification capabilities that automatically remove or mask personally identifiable information (PII), such as names, email addresses, and account information. This reduces the link between data and specific individuals or enterprises and helps prevent sensitive information from being leaked to third parties or learned by public models.

Enterprises can manage and export the data generated by their employees on the AnyInsight.ai platform.

Administrators can download user data in .csv, Markdown, and JSON formats once a month, covering the previous month’s data. A full export of all data may be performed within six months after a subscription is canceled.

Yes. AnyInsight.ai’s generative AI firewall is designed with contextual awareness and semantic analysis capabilities. It can identify not only explicit sensitive terms, but also synonyms, variant or obfuscated wording, and contextual intent in order to assess and block potential risks.

Its primary protection scope includes:

Prompt injection attacks: Blocking malicious instructions hidden within user inputs.
Sensitive data leakage: Preventing personal data (PII) or confidential business information from being transmitted or generated.
Non-compliant or high-risk content: Reducing the risk of generating inappropriate or policy-violating content.
Evasion attempts: Detecting attempts to bypass rules through paraphrasing, synonyms, or altered wording.

Real-Time Harmful Content Detection & Blocking

When the system detects abnormal instructions, sensitive personal information, hate speech, or other content that violates organizational governance policies, the submitted message may be blocked.

This is intended to protect data security, support regulatory and organizational compliance, and reduce potential risks.

Content that may trigger blocking includes general high-risk or inappropriate content, specific keywords configured by the organization’s administrator, and other restrictions defined within governance settings.Relevant risk information may be logged by the system and made available to administrators through the organization management console for further review and handling.

If you see a message stating that access has been denied due to governance policy, it means your organization’s administrator has applied security governance restrictions based on factors such as network IP address, login account, operating system, browser, or geographic location.

If you have any questions about login access or governance policy, you may take the following actions:

Contact your organization administrator: Ask them to review and resolve the login access issue.

AnyInsight.ai adopts architectures and management mechanisms aligned with international security best practices. It is designed with a Zero Trust Security Framework as its foundation and implements multi-layered protection measures to safeguard AI knowledge and user data during transmission, storage, and access. The platform is expected to obtain ISO 27001 certification by August 2026.

In terms of privacy and regulatory compliance, the platform also aligns with regulatory requirements such as the EU AI Act and the GDPR (General Data Protection Regulation), and continues to strengthen its personal data protection and AI governance mechanisms.

On the AnyInsight.ai platform, if a large language model experiences delayed responses, disconnections, or becomes unresponsive, it may be related to one of the following factors:

(1) Language model provider connection issues

If the third-party language model provider experiences service disruptions, connection failures, or high computational load, the model may respond more slowly, become temporarily unavailable, or disconnect.

We recommend trying again later or switching to another language model.

(2) User-side network connection interruption

If your network connection is unstable, temporarily interrupted, or your device or network environment encounters issues, message delivery may fail, responses may be interrupted, or the connection to the model may be terminated.

We recommend checking your current network connection.

(3) High network latency

If network latency is high, message transmission and model response times may be prolonged, which may result in slow responses, longer loading times, or what appears to be a disconnection.

Since each message must first go through data governance checks and encryption protection procedures, response time may also be slightly affected under certain circumstances.

We recommend checking your network quality and, if necessary, switching to another network environment before trying again.

(4) Impact of data governance checks and security protection mechanisms

On the AnyInsight.ai platform, every message is first processed through a generative AI firewall for data governance checks and encryption protection, which may affect response time.

In addition, if the task is more complex, the selected large language model is under heavy computational load, or the user’s network connection is unstable, responses may become slower or temporarily unresponsive.

To improve overall response efficiency, we recommend simplifying your input, breaking complex requests into smaller parts, and checking your network connection.

If this issue continues, you may use the following support channels:

AI Helper: Located at the upper-right corner of the page. You can communicate with the AI assistant directly from the message page.
Discord Community: On the AI Helper homepage, click “Join Discord Community” to report the issue.
Customer Support Email: customer_support@anyinsight.ai

Some large language models (LLMs), Agents, Prompt Templates, and MAIA modes may be restricted depending on your subscription plan and the permissions configured by your organization administrator.

As a result, some users may not be able to access certain features. We recommend first checking the permissions currently assigned to your account.

If you still cannot use the feature, you may take the following actions:

Contact your organization administrator: Ask them to verify and enable the required permissions.
AI Helper: Located at the upper-right corner of the page. You can communicate with the AI assistant directly from the message page.
Discord Community: On the AI Helper homepage, click “Join Discord Community” to report the issue.
Customer Support Email:customer_support@anyinsight.ai

MAIA (Multi-AI Agent) is AnyInsight’s core multi-AI collaboration mode. Based on the user’s task instructions and supporting materials, MAIA enables language models and agents with different capabilities to complement and cross-validate one another.

This helps improve the accuracy of complex task execution while reducing the likelihood of hallucinations.
Depending on task complexity, MAIA uses the following three logical processing modes:

Parallel Mode: Multiple models or agents respond to the same question simultaneously, helping users verify whether hallucinations or inconsistencies exist.

Integrative Mode: Responses from different models or agents are consolidated and compared, helping users broaden their perspectives and extend the depth of analysis.

Critique Mode: Models or agents review and critique one another’s responses, helping users identify different arguments, reasoning gaps, and points of divergence.

Not all AI agents enable MAIA mode by default.

Users can either allow MAIA to adjust automatically based on task complexity or configure it manually.

There are two main ways to use MAIA: directly from the chat input interface, or through collaboration between an Agent and MAIA.

Scenario 1: Enable MAIA Collaboration from the Chat Input Box

Step 1. Select a primary model in the chat interface. Then, hover over the chat selector to expand the conversation mode options.

Step 2. After expanding the options, select the MAIA mode you want to use: Parallel Mode, Integrative Mode, or Critique Mode.

Step 3. Select a second large language model (LLM) to add to the conversation. Once you send your message, MAIA mode will begin.

Scenario 2: Enable MAIA Collaboration with an Agent

Step 1. From the left sidebar on the homepage, navigate to Agent Studio. Then, select Agent Builder, or use Agent Controller to enter the edit page of the Agent you wish to enable MAIA for.

Step 2. In the Agent skills settings, enable MAIA Mode.

Step 3. MAIA mode will then be available within that Agent.

The core role of the second LLM is to validate, collaborate with, and detect discrepancies in the output of the primary model or Agent.

Therefore, model selection should be based on task requirements and resource efficiency, rather than simply choosing the most advanced model.

By uploading supporting documents or defining precise instructions through prompt engineering, users can strengthen the LLM’s understanding of domain-specific context and reduce performance differences between models. For ambiguous or exploratory tasks, using an advanced LLM for multi-round iterative collaboration remains a key approach to improving output quality.

AnyInsight.ai currently supports online credit card payments and provides real-time currency conversion for payments in USD, EUR, and TWD.

During registration and subscription, you may flexibly choose your preferred billing currency based on your needs. All transactions are processed through the third-party payment platform Stripe to ensure the security of your payment information.

AnyInsight.ai does not store your credit card information on its platform.

All sensitive payment information is encrypted and securely processed by Stripe, a trusted third-party payment platform, and stored within infrastructure designed to meet high security standards.

Stripe maintains security controls that comply with the Payment Card Industry Data Security Standard (PCI DSS), helping protect your payment information throughout the transaction process.

For more information, please refer to the following document:

Stripe Security Documentation: https://docs.stripe.com/security/guide

After completing payment, you can view your credit card payment information in the Manage Plan.

Step 1. Log in to AnyInsight.ai and go to the homepage.

Step 2. Click “Enter Org Administration”.

Step 3. In the left-side menu, go to “Settings” and select “Subscription”.

Step 4. On the “Subscription” page, click the “Manage Plan” button.

Step 5. You will be redirected to the Stripe page, where you can view your credit card information under “Payment method”.

Step 1. Log in to AnyInsight.ai and go to the homepage.

Step 2. Click “Enter Org Administration”.

Step 3. In the left-side menu, go to “Settings” and select “Subscription”.

Step 4. On the “Subscription” page, click the “Manage Plan” button.

Step 5. On the Stripe page, click “Invoice history”.

Step 6. You can then download the invoice and download the receipt.

If you have already subscribed to AnyInsight.ai, you may adjust the number of seats at any time and purchase additional seats flexibly based on your actual usage needs.

Step 1. From the left-side navigation bar on the homepage, click “Enter Org Administration” to enter the organization management page.

Step 2. On the organization management page, select “Settings” from the left-side menu, then click “Subscription” to be redirected to the third-party Stripe billing page where you can review your current plan.

Step 3. Click “Manage Plan” on the page to go to the plan and payment page, then click “Update Subscription” again.

Step 4. You may freely change your organization’s plan, adjust the number of Seats, and switch your billing cycle between Monthly and Yearly.

Additionally, if you choose to upgrade your plan, you will be able to access the services included in the higher-tier plan immediately after the upgrade.

The price difference between your current plan and the upgraded plan will be charged accordingly to ensure fair use and protect user rights.

If you choose to downgrade your plan, the downgrade will take effect automatically after your current plan period ends.

If you have other billing-related questions, you may use the following support channels:

AI Helper: Located at the upper-right corner of the page. You can communicate with the AI assistant directly from the message page.
Discord Community: On the AI Helper homepage, click “Join Discord Community” to send feedback.
Customer Support Email: customer_support@anyinsight.ai

If you do not yet have an AnyInsight.ai account, please first complete the standard user registration process and subscribe to a plan for the first company.

After that, log in to the platform, create a second organization, and subscribe to a plan for the second company.

If you already have an AnyInsight.ai account, you can simply log in, create a second organization, and complete the subscription for that organization.

(1) If you do not yet have an AnyInsight.ai account:

Step 1. Go to the AnyInsight.ai login page and click “Sign up”.

Step 2. Check the box to agree to the Terms of Use, then click “Next”.

Step 3. Check the box to agree to the Privacy Policy, then click “Next”.

Step 4. Enter your email address, click “Send Verification”, and check your inbox for the verification email.

(If you do not receive the email, please first check your spam folder. If you still have not received it, please try again after one minute.)

Step 5. After email verification is completed, the system will direct you to the account and company information page. Please fill in the information for the first company.

Step 6. After registration is complete, log in to your AnyInsight.ai account.

Step 7. Subscribe to a plan for the organization belonging to the first company.

Step 8. After the purchase is complete, click “Profile” in the left-side navigation bar on the homepage.

Step 9. Click “Switch Organization”. Once you enter the organization selection page, click “Create Organization”.

Step 10. Enter the information for the second company to create the second organization.

Step 11. Complete the subscription and payment process for the second organization.

(2) If you already have an account:

Step 1. After logging in to AnyInsight.ai, click “Profile” in the left-side navigation bar on the homepage.

Step 2. Click “Switch Organization”. Once you enter the organization selection page, click “Create Organization”.

Step 3. Enter the information for the second company to complete the creation of the second organization.

Step 4. Complete the subscription and payment process for the second organization.

Once the above steps are completed, you will have two separate organizations, each corresponding to a different company, and the two organizations will remain independent from one another.

You can review your current subscription plan, seat count, and billing cycle on the Manage Plan page to confirm whether the change has taken effect.

Step 1. From the left-side navigation bar on the homepage, click “Profile”.

Step 2. Click “Manage Plan” to enter the billing management page.

Step 3. On this page, you can view your current subscription plan, seat count, and billing cycle.

AnyInsight.ai utilizes an auto-renewal mechanism, so you don't have to worry about service interruptions due to plan expiration.

At the end of each billing cycle, the system will automatically process the payment and renew your current subscription.

You may adjust your subscription status at any time on the Manage Plan page. If you cancel your subscription, the service will remain available until the end of the current subscription period.

If you have other billing-related questions, you may use the following support channels:

AI Helper: Located at the upper-right corner of the page. You can communicate with the AI assistant directly from the message page.
Discord Community: On the AI Helper homepage, click “Join Discord Community” to send feedback.
Customer Support Email:customer_support@anyinsight.ai

AnyInsight.ai uses an automatic renewal mechanism. At the end of each billing cycle, the system will automatically charge your account and renew your subscription based on your current plan.

If you wish to stop future automatic charges, please go to the Manage Plan page and cancel renewal.

Step 1. Click “Enter Org Administration” from the left-side navigation bar.

Step 2. In the left-side menu, go to “Settings” and select “Subscription”.

Step 3. On the “Subscription” page, click the “Manage Plan” button.

Step 4. On the Stripe page, you can review your subscription details and click “Cancel subscription”.

If you have other billing-related questions, you may use the following support channels:

AI Helper: Located at the upper-right corner of the page. You can communicate with the AI assistant directly from the message page.
Discord Community: On the AI Helper homepage, click “Join Discord Community” to send feedback.
Customer Support Email: customer_support@anyinsight.ai

If you cancel your subscription before the end of the 30-day trial period, you will not be charged any fees.

Please note that once a subscription charge has been successfully processed, it is non-refundable. However, you may continue using the service until the end of the current billing period. We recommend reviewing the Terms of Service and pricing information carefully before making a purchase.

For more information, please refer to:

Terms of Service: https://www.anyinsight.ai/policies/terms-of-service

If you have other billing-related questions, you may use the following support channels:

AI Helper: Located at the upper-right corner of the page. You can communicate with the AI assistant directly from the message page.
Discord Community: On the AI Helper homepage, click “Join Discord Community” to send feedback.
Customer Support Email:customer_support@anyinsight.ai

The data retention period varies depending on the subscription plan.

Pro plan provide data retention for up to 1 year.
Growth and Enterprise plans provide data retention for up to 2 years and Customer.

Plan	Pro	Growth	Enterprise
Data Query Period Days	30 days	90 days	Customer
Data Retention Period	1 Year	2 Years	2 Years

Plan

Data Query Period Days

Pro

30 days

Growth

90 days

Enterprise

Customer

Plan

Data Retention Period

Pro

1 Year

Growth

2 Years

Enterprise

2 Years

AnyInsight.ai reserves the right to modify its service offerings at any time. The actual items and services provided shall be subject to the latest announcements on the platform.

You can access the “How to Build Prompt Templates” dialogue in the “Get Started” area on the homepage to learn about template creation, standards, and techniques.

Step1. Go to the “Get Started” area on the homepage.

Step2. Select the “How to Create Prompt Templates” dialogue.

You can also click “Prompt Tools” in the chat input box and choose a template to apply, which will insert the template content into the chat input.

Step1. Click “Prompt Tools” below the chat input box.

Step2. Select “Apply Template”.

Step3. Search for and select the prompt template you wish to use.

Step4. Review the template details and click “Apply”.

Step5. The template will be pasted into the chat box; simply add your specific prompt and send.

Before entering your message to the model, you can also click “Prompt Tools” in the conversation input box and select “Prompt Wizard”. This feature allows you to define your desired output and use AI to automatically generate a prompt template from your instructions for quick application.

Step 1. Click “Prompt Tools” below the conversation box.

Step 2. Select “Prompt Wizard”.

Step 3. Fill in the Role, Goal, Desired Output, and Constraints in Prompt Wizard.

For over 100+ scenarios including HR, Administration, and Marketing, you can install templates directly from the “Inventory”:

Step1. Select “Inventory” from the workspace on the left sidebar.

Step2. Choose “Prompt Template” from the menu.

Step3. Click on the desired template and select “Install”. Once installed, these templates will be available for use on your homepage and in the chat input box.

Step 1. On the left sidebar of the homepage, open the “Agent Studio” dropdown menu and click “Agent Builder” to enter the agent creation page.

Step 2. Complete the basic agent settings, including the agent name, description, language model, instructions, and other default conditions.

Step 3. Click “Save as a New”.

Step 4. Enter a name for the new preset and save it as a preset.

Step 5. You can then test the agent’s responses in the “Sandbox”.

Step 6. Click “Save and Publish” to create your dedicated agent.

You can use the “Upload Knowledge Base” feature on the agent creation page to provide dedicated data for the agent to retrieve and reference.

There are two ways to update an agent’s knowledge. You can update the knowledge while creating an agent in “Agent Builder”, or edit an existing agent directly through “Agent Controller”.

Updating Knowledge in Agent Builder

Step 1. On the left sidebar of the homepage, go to “Agent Studio” and enter the “Agent Builder” page.

Step 2. On the “Agent Builder” page, find the “Knowledge” section.

Step 3. Click the “Upload File” button to update the agent’s knowledge. See Note [1].

Step 4. Click the “Upload URL” button to import a URL. See Note [2].

Editing an Existing Agent in Agent Controller

Step 1. On the left sidebar of the homepage, go to “Agent Studio” and enter “Agent Controller”.

Step 2. In the right-side column of the agent you want to edit, click “Edit”.

Step 3. On the agent editing page, find the “Knowledge” section and click the “Upload File” button to update the agent’s knowledge. See Note [1].

Step 4. Click the “Upload URL” button to import a URL. See Note [2].

Notes

Note [1]: Supported file types include PDF, TXT, CSV, DOCX, XLSX, and HTML. Each file can be up to 10 MB, with a maximum of 20 files.

Note [2]: Up to 5 URLs are supported. URL content is automatically updated to the knowledge base weekly. Updates usually take effect within a few minutes; invalid sources will be skipped.

Yes. You can install Pre-built Agents from the Inventory in different Workspaces.

The platform currently provides more than 350 Pre-built Agents, covering various use cases such as human resources, administration, marketing, sales, customer service, and operations. You only need to select a suitable agent and complete the installation to start using it in the designated Workspace, without having to create an agent from scratch.

Step 1. On the left sidebar of the homepage, expand Inventory under the corresponding Workspace.

Step 2. After expanding the menu, click “Agent” to enter the Pre-built Agents page. You can browse agents by category and select the agent you want to install.

Step 3. Select an agent and click “Install”.

Step 4. After the installation is successful, the agent will appear in the default “Workspace” you entered.

To learn more about AnyInsight.ai’s agent features and use cases, please refer to the following page:

Link: https://www.anyinsight.ai/pages/agent

AnyInsight.ai currently offers three subscription plans: Pro, Growth, and Enterprise, designed to meet the needs of teams and organizations of different sizes. Note [1]

Users can choose the most suitable plan based on their organization size, collaboration needs, security requirements, and intended use cases.

Plan	Suitable For
Pro	Suitable for freelancers, small businesses, or content creation teams of 1–100 users. Provides higher AI usage quotas and advanced features to help individuals or teams improve daily work efficiency.
Growth	Designed for organizations of 1–999 users, suitable for growing mid-sized companies and cross-functional teams. Provides more comprehensive enterprise-grade security protection and team collaboration features to help businesses integrate AI into daily operations.
Enterprise	Suitable for large enterprises, groups, or organizations. Provides customized plans, advanced permission management, enterprise-grade deployment support, and dedicated customer service to meet higher standards for security, governance, and custom integration needs.

Plan

Pro

Suitable For

Suitable for freelancers, small businesses, or content creation teams of 1–100 users. Provides higher AI usage quotas and advanced features to help individuals or teams improve daily work efficiency.

Plan

Growth

Suitable For

Designed for organizations of 1–999 users, suitable for growing mid-sized companies and cross-functional teams. Provides more comprehensive enterprise-grade security protection and team collaboration features to help businesses integrate AI into daily operations.

Plan

Enterprise

Suitable For

Suitable for large enterprises, groups, or organizations. Provides customized plans, advanced permission management, enterprise-grade deployment support, and dedicated customer service to meet higher standards for security, governance, and custom integration needs.

Note [1]: As technology advances and product services continue to evolve, AnyInsight.ai may adjust and optimize the features, model support, usage quotas, and service details included in each subscription plan from time to time. The actual available features and limitations shall be based on the latest information listed on the official website, in-platform announcements, or subscription page.

	Pro	Growth	Enterprise
Seats	1~100	1~999	Unlimited
Chat MAIA Mode	✓	✓	✓
Agent MAIA Mode	✓	✓	✓
Language Model Support	20+	20+	20+ · BYO cloud-native LLM model*
Pre-built Agents	At least 50	All agents (350+)	All agents (350+) + Custom Build*
Custom Agents	Default and Personal Workspaces	Cross-workspace*	Cross-workspace*
Prompt Tools and Prompt Templates	✓	✓	✓
Connectors	100 essentials	All connectors (500+)	All connectors (500+)
Multilingual Support (9 Languages) Note [3]	✓	✓	✓
Customer Support	AI Helper, Email, Discord	Priority Support*	Dedicated Customer Support*
Walled Workspace	Default and Personal workspace	✓ Unlimited	✓ + Cross-organization support*
Identity & Access Management	Standard	Full RBAC + SSO* + Passkey*	RBAC + SSO* + SCIM**
Dashboard	Shared Workspace	✓	✓
Inventory Hub	—	✓	✓
Dialogue logs	Shared Workspace	✓	✓
System Log Retention	—	✓ 90 days	✓ 6 months
Connector Logs	—	✓ 90 days	✓ 6 months + SIEM* integration
Usage Analytics	—	✓	✓ + Custom Reports*
Data Vault	—	✓	✓
Firewall	—	Standard Rules	Standard Rules + Custom Rules
AI Preventive Access Control	—	✓	Custom Rules
Personally Identifiable Information (PII) Filtering	—	✓	✓ + Custom Rules
Prompt injection	—	✓	✓ + Custom Rules
Proactive Risk Monitoring	—	✓	✓ + Advanced Alerts
Alert Notifications	Email	Email	Email + Webhook + SIEM**
Data Protection Zone Data Retention	✓ 30 days	✓ 90 days	✓ Custom Retention
ISO 27001 / GDPR / EU AI Act	Compliance	Compliance	Compliance

Seats

Pro

1~100

Growth

1~999

Enterprise

Unlimited

Chat MAIA Mode

Pro

✓

Growth

✓

Enterprise

✓

Agent MAIA Mode

Pro

✓

Growth

✓

Enterprise

✓

Language Model Support

Pro

20+

Growth

20+

Enterprise

20+ · BYO cloud-native LLM model*

Pre-built Agents

Pro

At least 50

Growth

All agents (350+)

Enterprise

All agents (350+) + Custom Build*

Custom Agents

Pro

Default and Personal Workspaces

Growth

Cross-workspace*

Enterprise

Cross-workspace*

Prompt Tools and Prompt Templates

Pro

✓

Growth

✓

Enterprise

✓

Connectors

Pro

100 essentials

Growth

All connectors (500+)

Enterprise

All connectors (500+)

Multilingual Support (9 Languages) Note [3]

Pro

✓

Growth

✓

Enterprise

✓

Customer Support

Pro

AI Helper, Email, Discord

Growth

Priority Support*

Enterprise

Dedicated Customer Support*

Walled Workspace

Pro

Default and Personal workspace

Growth

✓ Unlimited

Enterprise

✓ + Cross-organization support*

Identity & Access Management

Pro

Standard

Growth

Full RBAC + SSO* + Passkey*

Enterprise

RBAC + SSO* + SCIM**

Dashboard

Pro

Shared Workspace

Growth

✓

Enterprise

✓

Inventory Hub

Pro

—

Growth

✓

Enterprise

✓

Dialogue logs

Pro

Shared Workspace

Growth

✓

Enterprise

✓

System Log Retention

Pro

—

Growth

✓ 90 days

Enterprise

✓ 6 months

Connector Logs

Pro

—

Growth

✓ 90 days

Enterprise

✓ 6 months + SIEM* integration

Usage Analytics

Pro

—

Growth

✓

Enterprise

✓ + Custom Reports*

Data Vault

Pro

—

Growth

✓

Enterprise

✓

Firewall

Pro

—

Growth

Standard Rules

Enterprise

Standard Rules + Custom Rules

AI Preventive Access Control

Pro

—

Growth

✓

Enterprise

Custom Rules

Personally Identifiable Information (PII) Filtering

Pro

—

Growth

✓

Enterprise

✓ + Custom Rules

Prompt injection

Pro

—

Growth

✓

Enterprise

✓ + Custom Rules

Proactive Risk Monitoring

Pro

—

Growth

✓

Enterprise

✓ + Advanced Alerts

Alert Notifications

Pro

Growth

Enterprise

Email + Webhook + SIEM**

Data Protection Zone Data Retention

Pro

✓ 30 days

Growth

✓ 90 days

Enterprise

✓ Custom Retention

ISO 27001 / GDPR / EU AI Act

Pro

Compliance

Growth

Compliance

Enterprise

Compliance

Note [2]: AnyInsight.ai reserves the right to adjust its service offerings at any time. The actual available items shall be based on the latest platform announcements.

Note [3]: Multilingual support includes: English, Chinese, Thai, Spanish, French, Portuguese, German, Dutch, and Italian.

Note [4]: As technology advances and product services continue to evolve, AnyInsight.ai may adjust and optimize the features, model support, usage quotas, and service details included in each subscription plan from time to time. The actual available features and limitations shall be based on the latest information listed on the official website, in-platform announcements, or subscription page.

Note [5]: * indicates that the feature is expected to launch in Q3 2026; ** indicates that the feature is expected to launch in Q4 2026.

The table below describes the language model usage quotas and image generation limits for each AnyInsight.ai subscription plan. Unless otherwise specified, all quotas are calculated on a monthly basis and are counted per seat. Note [6]

	Pro	Growth	Enterprise
Monthly Usage for Base Language Models	3,000 per seat	5,000 per seat	Unlimited
Monthly Usage for Advanced Language Models Note [6], Note [7]	3,000 uses per seat	5,000 per seat	Unlimited
Monthly Image Generation Limit	300 images per seat	500 images per seat	500 images per seat
Monthly Nano Banana Pro Image Generation Limit	—	50 images per seat	100 images per seat

Monthly Usage for Base Language Models

Pro

3,000 per seat

Growth

5,000 per seat

Enterprise

Unlimited

Monthly Usage for Advanced Language Models Note [6], Note [7]

Pro

3,000 uses per seat

Growth

5,000 per seat

Enterprise

Unlimited

Monthly Image Generation Limit

Pro

300 images per seat

Growth

500 images per seat

Enterprise

500 images per seat

Monthly Nano Banana Pro Image Generation Limit

Pro

—

Growth

50 images per seat

Enterprise

100 images per seat

Usage Calculation Rules Note [8]

Billing Cycle: All usage quotas are calculated on a monthly basis.
MAIA Usage Calculation: When using MAIA features, each operation will be counted as 2 uses.
Nano Banana Pro Image Generation Limit: Images generated using Nano Banana Pro will be counted toward the monthly image generation limit.

Note [6]: Advanced language models include: GPT 5.4, GPT 5.1, GPT 5, Gemini 2.5 Pro, Gemini 3.0 Pro, Gemini 3.1 Pro, Nano Banana v2, Nano Banana, Nano Banana Pro, Claude Sonnet 4.5, and Claude Opus 4.5. For more details, please refer to 4.Which base and advanced large language models does the AnyInsight.ai platform support?

Note [7]: The models supported by AnyInsight.ai may change depending on updates, deprecations, or service policy adjustments made by each model provider.

Note [8]: As technology advances and product services continue to evolve, AnyInsight.ai may adjust and optimize the features, model support, usage quotas, and service details included in each subscription plan from time to time. The actual available features and limitations shall be based on the latest information listed on the official website, in-platform announcements, or subscription page.

All AnyInsight.ai plans provide access to both Base Large Language Models and Advanced Large Language Models. Note [9], Note [10]

Base Large Language Models	Advanced Large Language Models
GPT 4o GPT 4o mini GPT 4.1 Gemini 2.0 Flash Claude Haiku 4.5 Mistral Large TAIDE Perplexity Grok 4 GPT OSS 120b GPT OSS 20b	GPT 5.4 GPT 5.1 GPT 5 Gemini 2.5 Pro Gemini 3.0 Pro Gemini 3.1 Pro Nano Banana v2 Nano Banana Nano Banana Pro Claude Sonnet 4.5 Claude Opus 4.5

Base Large Language Models

GPT 4o
GPT 4o mini
GPT 4.1
Gemini 2.0 Flash
Claude Haiku 4.5
Mistral Large
TAIDE
Perplexity
Grok 4
GPT OSS 120b
GPT OSS 20b

Advanced Large Language Models

GPT 5.4

GPT 5.1

GPT 5

Gemini 2.5 Pro

Gemini 3.0 Pro

Gemini 3.1 Pro

Nano Banana v2

Nano Banana

Nano Banana Pro

Claude Sonnet 4.5

Claude Opus 4.5

Note [9]: The models supported by AnyInsight.ai may change depending on updates, deprecations, or service policy adjustments made by each model provider.

Note [10]: As technology advances and product services continue to evolve, AnyInsight.ai may adjust and optimize the features, model support, usage quotas, and service details included in each subscription plan from time to time. The actual available features and limitations shall be based on the latest information listed on the official website, in-platform announcements, or subscription page.

AnyInsight.ai currently focuses on a fully cloud-based SaaS solution and does not offer on-premises deployment at this time. Enterprises can quickly adopt generative AI applications without having to deploy models themselves or build complex environments.

The cloud platform offers strong scalability, ready-to-use access, and centralized management, effectively lowering the barrier to adoption while meeting enterprise needs across different stages of growth.

If you have project customization needs or require volume purchasing discounts, we recommend contacting our customer support team at customer_support@anyinsight.ai.

We will evaluate the most suitable plan based on your company size, use cases, and implementation requirements.

For further assistance, you may also fill out the form through the link below. Our team will contact you as soon as possible to help your organization smoothly adopt AI.

Customer Support Email: customer_support@anyinsight.ai
Contact Form: https://www.anyinsight.ai/zh/pages/aboutus
AI Helper: Located at the upper-right corner of the page. You can communicate with the AI assistant directly from the message page.
Discord Community: On the AI Helper homepage, click “Join Discord Community” to report the issue.

You can click Forgot Password on the login page to reset your password.

Step 1. Go to the AnyInsight.ai login page and click “Forgot Password”.

Step 2. Enter the email address used for registration. The system will send a verification link to that email address. Click the link to be redirected to the password reset page.

If you do not receive the email, please check your spam folder or contact the organization administrator of your account.

Step 1. Log in as an “Org Admin” or “Org Owner”. On the left sidebar, click Enter “Org Administration”.

Step 2. In the “Admin Panel” section of the left menu, go to “User Groups”, then select the “User” page.

Step 3. On the “User” page, the administrator can view all users in the organization. Find the relevant user, click “Edit User”, and select “Reset Password”.

Step 4. The system will send a password reset email to the email address registered with that account.

Yes. The platform supports logging in to the same account on multiple devices or browser tabs at the same time.

However, to ensure the best user experience and consistent data synchronization, we recommend avoiding making configuration changes on multiple devices at the same time.

Repeated verification failures may prevent users from logging in successfully. Possible causes include the following:

(1) Not receiving the email:

Please check whether the verification email is in your spam folder. If not, wait for one minute and try again.

(2) Expired verification email:

The user may be trying to enter a verification code from an expired email.

If you still cannot log in, please contact your organization administrator or report the issue to AnyInsight.ai customer support at customer_support@anyinsight.ai.

The Inventory Controller is the central panel where administrators can manage AI resources across the organization. From this page, you can view and manage the enabled status of all Agents, Prompt Templates, and Connectors throughout the organization, supporting resource allocation and access control audits.

To control access to specific resources, go to the Inventory Controller and disable the relevant items:

Effect of disabling a resource: When an Agent, Prompt Template, or Connector is disabled, other users in the organization will no longer be able to view, select, or use that resource in their respective Workspaces.

Steps:

Step 1. Go to the home page and click Organization in the left-hand navigation menu to open the Organization Administration page.

Step 2. Under the Settings section near the bottom of the page, click Inventory Controller.

Step 3. Use the tabs at the top of the page to switch between Agents, Prompt Templates, and Connectors. Locate the item you want to manage, then switch off the status toggle on the far right.

To review all plan details associated with your organization account, go to the Subscription page:

Plan Overview: Review your current subscription tier, the number of licenses currently in use, and the total license limit.
Core Models: View the complete list of large language models (LLMs) supported and available under your current plan.
Data and Privacy: Review data access permissions and retention periods, and export data from this page.

Steps:

Step 1. Go to the home page and click Organization in the left-hand navigation menu to open the Organization Administration page.

Step 2. Under the Settings section near the bottom of the page, click Subscription.

Step 3. On the Subscription page, you can review the subscription overview, supported model list, and other plan information.

Users with organization administration permissions can update the company's basic information by following these steps:

Steps:

Step 1. Go to the home page and click Organization in the left-hand navigation menu to open the Organization Administration page.

Step 2. Under the Settings section near the bottom of the page, click Org Info.

Step 3. In the General section, click and edit the organization name and business registration number fields.

Step 4. After entering the updated information, click Save. The system will save the revised organization information.

To strengthen your organization's brand identity within the platform, the system supports brand customization:

Upload requirements and recommendations:

Wordmark: A ratio of approximately 3:1 and a minimum size of 360 x 120 px are recommended. Because the image will be displayed in both light and dark themes, select an image that remains clear against both backgrounds. Supported formats include PNG, JPEG, SVG, and WebP. The file size must not exceed 2 MB.
Icon: A square image with a minimum size of 128 x 128 px is recommended. Supported formats include PNG, JPEG, SVG, and WebP. The file size must not exceed 2 MB.

To help administrators stay informed about the organization's operational status, such as critical errors, unusual usage activity, or policy violations, you can enable system notifications:

Notification Settings: Once enabled, the system does not send a separate email for every event. Instead, it consolidates events based on their severity and frequency, then sends a notification at the interval you specify. This helps prevent excessive emails from creating unnecessary administrative overhead.

Steps:

Step 1. Go to the home page and click Organization in the left-hand navigation menu to open the Organization Administration page.

Step 2. Under the Settings section near the bottom of the page, click Org Info.

Step 3. Scroll to the Notification Settings section at the bottom of the page, then switch on Send Email Notifications.

Step 4. Configure the Send Interval to determine how frequently the system consolidates events and sends notifications to administrators.

The Dashboard provides organization administrators with an organization-wide view of platform usage through visual charts. It is divided into four main sections: Key Statistics, User Behavior, User Environment Overview, and Security Analysis. These sections help you monitor AI resource consumption, team usage patterns, and potential security risks.

You can adjust the reporting period at any time to review performance or investigate anomalies within a specific time range:

Steps:

Step 1. Locate the time selector in the upper-right corner of the Dashboard page.

Step 2. Select a shortcut option, such as 1D, 7D, 30D, 3M, or 6M, to quickly apply a commonly used time range.

Step 3. To specify a custom date range, click the calendar icon and select the desired start and end dates.

This tab provides an overview of AI resource usage across the organization, including:

Core Metrics: Quickly review upward or downward trends in Active Users, Active Agents, Active LLMs, and Dialogue Count.
Rankings and Distribution Charts: View the Active Users List, Active Agents List, LLM Usage distribution, and request-based charts such as Users By Requests and Agents By Requests. These insights help you identify the resources used most frequently within the organization.

To analyze team interactions, switch to the User Behavior tab in the Dashboard. This tab provides detailed interaction data, including:

Activity and Sign-In Analysis: Review the User Activity Time Breakdown and Number of Logins.
Session Status: Track Average Session Time and Average Idle Time.
Activity Types: Use Content Category Distribution, such as the proportions of interacting, reading, and idle activity, to understand how members use and collaborate with AI.

Yes. Go to the User Environment Overview tab. This section supports basic IT environment inventory and sign-in anomaly auditing:

Device and Software Distribution: Pie charts record and analyze users' Device Type, OS, and Browsers.
Geographic Distribution: Active Users By City and Active Users by Location display the distribution of active users across cities and regions through bar charts and a map.

Security Analysis is an important component of organizational data protection and auditing. It helps administrators monitor and analyze abnormal events that trigger data loss prevention (DLP) or Firewall policies. This tab provides the following sections for a comprehensive security assessment:

Organization Security Status: Review Prompts Scanned and quickly monitor the total number of Blocked Events, Masked Events, and Alerted Events, along with their increases or decreases.
Organization Violation Distribution: Use the Organization Violation Details Distribution and Organization Violation Themes Distribution pie charts to identify the security risk categories most frequently triggered by the team, helping you adjust security policies accordingly.
High-Risk User Tracking: Use the Violation Details per High-Risk User and Violation Themes per High-Risk User charts to identify the users who trigger violations most frequently and their corresponding risk categories. This allows administrators to conduct internal awareness training, investigate incidents, or adjust access controls.
Violation Event Trends: Use the Governance Events Trend line chart to track the frequency of security events within a selected time range, helping you assess the organization's recent security posture and the effectiveness of its protection policies.

These events represent different levels of system response to potentially risky content:

Blocked: The system directly intercepts and prevents a high-risk or severely non-compliant prompt from being submitted.
Masked: The system allows the prompt to be submitted but automatically replaces sensitive information, such as phone numbers or addresses, with asterisks or other designated characters to prevent confidential information from being exposed.
Alerted: The system allows the content to be submitted without masking it. However, because the content may present a potential risk, the event is flagged and recorded for subsequent administrator review.

Yes. To focus on specific metrics, click Layout in the upper-right corner of the page. You can adjust the chart arrangement and displayed sections according to your viewing preferences and management priorities, allowing you to create a data-monitoring layout that best suits your needs.

Dashboard data is continuously aggregated in the background. To ensure that you are viewing the latest available data or to obtain more information about a metric:

Manual Refresh: Click the Refresh icon in the upper-right corner of the page, next to Layout, to retrieve and load the latest data.
Tooltips: If the definition of a chart or metric is unclear, hover over the "?" icon next to the chart title. The system will display a detailed definition and explanation of the metric.

If the Dashboard does not display data, it is usually due to one of the following common reasons:

No Activity Within the Selected Time Range: Check the time selector in the upper-right corner and confirm whether organization members signed in to the platform or interacted with AI during the selected date range.
Data Is Still Being Accumulated: If the organization was created recently or new members were only just invited, the system may need time to accumulate interaction records and generate statistical charts. Review the Dashboard again after members have started using the platform.

The System Log provides a comprehensive security and administrative audit trail for the organization. It records system activities performed by all organization members, including but not limited to user sign-in and data access activities, workspace switching, organization information views, permission changes, and knowledge base updates. This allows administrators to conduct compliance audits at any time and determine who performed a particular action and when it occurred.

You can quickly locate the relevant records using the search and filtering features in the System Log:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Log section in the left-hand navigation menu, click System Log.

Step 3. Enter the user name in the search bar at the top of the page. You can also use the time selector on the right to narrow the search range.

Step 4. Locate the relevant record in the list and click View on the far right.

Step 5. A details panel will open on the right, displaying complete operator information for the activity, including the source IP address, client location, operating system, and browser version.

Yes. The Dialogue Log consolidates AI interaction histories from all workspaces within the organization, helping administrators ensure that AI resources are used appropriately:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Log section in the left-hand navigation menu, open Dialogue Log. The list provides an overview of all conversation titles, the LLMs used, and the number of uploaded and output files.

Step 3. Locate the conversation record you want to review and click View on the far right.

Step 4. The Detailed Conversation Record window will open on the right. The left side displays the user's prompts and the complete text and image content returned by the AI agent, while the right side displays the conversation's basic information and execution details.

The Dialogue Log incorporates detection results from data loss prevention (DLP), allowing you to identify high-risk conversations easily:

Statistics Overview: Above the list, the system displays the total number of Alerts, Blocks, and Masks within the selected time range, helping you quickly identify high-risk conversations.
List View: Check the Alerts, Blocks, and Masks columns on the right side of the list. If a conversation triggered a security rule, the corresponding column displays the number of times it was triggered. After clicking View, you can review the blocking details in the Firewall section in the lower-right corner.

The Automation Log is specifically used to track the execution history of all automation tasks within the organization, rather than individual user-AI conversations. When a user configures a sequence of AI workflow steps, each task run, the agents executed, and the language models used are recorded separately in the Automation Log for administrator review and auditing.

You can review the detailed task record to examine each execution step:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Log section in the left-hand navigation menu, open Automation Log, then use the search bar at the top to filter records by task name.

Step 3. Locate the relevant execution record and click View on the far right.

Step 4. In the Detailed Task Record window, the left side lists each execution step in sequence, such as Step 1, followed by the final Summary. The right-side panel displays the connectors, uploaded files, and output results used in each step, helping you quickly identify the step where the workflow failed or was interrupted.

The Shared Conversation Log is an important auditing tool for preventing the improper disclosure of internal company information. When an organization member creates a share link for a conversation with an AI and shares it with other users within the organization, the system creates a corresponding record here. The list clearly identifies the Sharer, the sharer's email address, the shared conversation title, and the sharing time.

Yes. The system records access activity for each link:

Quick List View: In the list, you can directly review the link's View Count and Last View time.
Detailed Access Records: Click View on the far right of the list to open the Access Records panel on the right. This panel displays detailed viewer information and the status of related files, helping you assess the potential scope of data exposure.

The Connector Log is designed to monitor interactions between AI agents and third-party tools, including external systems. When an AI agent calls an external tool, all connection requests, execution times, and statuses are recorded here. You can switch between View by Prompt and View by Action in the upper-left corner to track API call records.

Yes. The Connector Log provides developer-level debugging information:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Log section in the left-hand navigation menu, open Connector Log. Locate the connection record you want to debug and click View.

Step 3. The Connector Execution Record window will open on the right.

Step 4. In the right-side panel, you can expand and review the Input Parameters sent by the system and the Output Results returned by the external system. The data is displayed in standard JSON format, including fields such as message, error, and error codes, allowing technical personnel to copy the data directly and troubleshoot issues efficiently.

You can use the workspace list to review the resources across your organization:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Workspace.

Step 3. On this page, you can view a list of all workspaces, including their names, types, creators, and total number of members.

Step 4. To find a specific workspace, enter the workspace name in the search bar at the top of the page, or use the adjacent Filter button to perform an advanced search.

You can rename a workspace at any time to better suit your team's needs:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Workspace.

Step 3. Locate the target workspace in the workspace list and click Edit on the far right.

Step 4. An editing window will appear. Enter the new workspace name in the input field. Workspace names can contain up to 20 characters. After confirming that the information is correct, click Save.

If a workspace is temporarily no longer needed or needs to be made available again, you can change its status at any time:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Workspace.

Step 3. Locate the target workspace in the workspace list, then use the Status toggle on the right side of the row to enable or disable it.

You can switch between individual users and user groups in the Admin Panel:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click User Groups.

Step 3. Use the tabs in the upper-right corner of the page. Select Users to view a list of all members and their account statuses, or select User Groups to view the groups created within the organization.

Grouping specific users makes it easier to assign permissions later:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click User Groups.

Step 3. Switch to the User Groups tab in the upper-right corner of the page, then click Create User Group in the upper-left corner.

Step 4. In the group editing window, enter a user group name, then click Next.

Step 5. Select the users you want to add to the group from the list below. You can also use the search bar to quickly find specific users. Click Save to create the group.

You can invite new members by entering their information manually or importing them in bulk:

Steps for manual entry:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click User Groups.

Step 3. Switch to the Users tab, then click the orange Invite to Organization button in the upper-left corner.

Step 4. In the window that appears, enter the user's user name and email address.

Step 5. Click Next and assign the user a role within the organization.

Step 6. Confirm that the user's information is correct, then complete the invitation process.

Steps for importing users with a .CSV file:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click User Groups.

Step 3. Switch to the Users tab, then click Invite to Organization in the upper-left corner.

Step 4. In the window that appears, click Download Template to download the .csv file.

Step 5. Enter the members' names and email addresses in the file, then click Upload File to import the users in bulk.

Step 6. Click Next and assign the users their roles within the organization.

Step 7. Confirm that the users' information is correct, then click Invite to Organization to complete the invitation process.

To manage the account status of an existing member, follow these steps:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click User Groups.

Step 3. Make sure you are viewing the Users list. Locate the target user, then click Edit on the far right of the list.

Step 4. From the expanded drop-down menu, select the appropriate action:

Edit User: Updates the member's basic name information.
Disable: Temporarily suspends the user's access to the account.
Enable: Restores sign-in access for a disabled user. This option is available only when the user is currently disabled.
Reset Password: Sends the user a password reset notification.

You can click the Organization Topology button to quickly review the organization's members and their corresponding roles in a visual card-based view:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click User Groups.

Step 3. Click Open topology in the upper-right corner of the page.

Step 4. The Organization topology window will appear. From here, you can visually review cards for all members, along with their organizational levels and permission labels.

You can use the Privilege Manager to flexibly adjust the roles assigned to organization members at different scopes, such as the organization or workspace level:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Privilege Manager.

Step 3. Make sure the page is currently in Role Assignment mode in the upper-right corner.

Step 4. On the User role assignment tab, locate the target user or group and click Edit on the far right.

Step 5. In the window that appears, click + Add Role.

Step 6. Select the applicable Scope and corresponding Role from the drop-down menus.

Step 7. After completing the configuration, click Confirm to apply the new permissions.

If the system's default roles do not fully meet your team's needs, you can create a custom role and define the specific features and actions that the role can access:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Privilege Manager.

Step 3. Click Role Management in the upper-right corner of the page to switch to that mode.

Step 4. Click + Add Role in the left-hand panel.

Step 5. On the Create New Role page, enter a Role Name, select the applicable Scope—organization or workspace—and optionally enter a Description.

Step 6. In the Permissions section below, expand each category as needed and select the specific features or actions that the role is permitted to access or perform.

Step 7. After confirming that all permissions are configured correctly, click Create Role at the top of the page.

Administrators can centrally edit and maintain agent settings within the organization. Note that any changes made here will directly affect the agent status and settings for all users in the organization:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Agents.

Step 3. Locate the target agent in the agent list, then click Edit under the Actions column on the far right—the pencil icon.

Step 4. Make the required changes on the editing page and save them. The changes will then be applied across the entire organization.

If certain agents are temporarily not needed by the team, you can disable them. Once an agent is disabled, all users within the organization will be unable to view or use it:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Agents.

Step 3. Locate the target agent in the agent list, then use the toggle in the Enabled column to enable or disable it.

You can centrally update prompt content through the Organization Management panel. After the changes are saved, all users in the organization will receive the latest version whenever they apply the template:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Prompt Templates.

Step 3. Locate the template you want to modify in the list, then click Edit on the far right.

Step 4. Update the template name, description, or prompt content, then click Save to apply the changes.

If a prompt template is no longer applicable or needs to be temporarily removed from use, you can manage access to it using the toggle. When the template is disabled, users within the organization will no longer be able to select it:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Admin Panel section in the left-hand navigation menu, click Prompt Templates.

Step 3. Locate the target template in the list, then use the toggle in the Enabled column to enable or disable it.

The Data Vault is the centralized control center for all AI-related file assets within an organization. Files uploaded directly by users and documents shared in conversations are consolidated and managed here. This panel allows administrators to review each file's source, type, and size, as well as track its security scan status.

You can use the list to review detailed information about each file and its source.

Basic Information: Includes the upload time, file name, file type (such as CSV, PPTX, PDF, or JPG), and file size.
Source and Owner: Clearly identifies the user name, email address, source (such as direct upload, conversation, or AI-generated content), and the specific source name.

You can quickly assess a file by checking the Governance Status column in the list:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Security section in the left-hand navigation menu, click Data Vault.

Step 3. Check the Governance Status column. The system uses color-coded labels to display the file's status, such as Safe, Unchecked, or Masked, helping you quickly identify files that may pose potential security risks.

The Firewall serves as the first line of defense for protecting organizational data. The Policy Management tab provides multiple categories of protection:

Access Control: Restricts sign-in and platform activities based on browser, device type, operating system, time range, or IP address.
Content Filter: Filters files or prompts based on keywords or semantic topics.

Sensitive Info.: Automatically detects and protects financial information, health records, personally identifiable information (PII), cloud credentials, and other sensitive data.
Toxic Content: Protects against inappropriate content involving harassment, hate, self-harm, and similar categories.

You can flexibly adjust a policy's status at any time based on your organization's security requirements:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Security section in the left-hand navigation menu, click Firewall.

Step 3. Make sure the Policy Management tab at the top of the page is selected. Locate the target policy in the list, then use the Status toggle on the far left to enable or disable it.

You can use the Content Filter feature to create customized keyword protection rules:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Security section in the left-hand navigation menu, click Firewall.

Step 3. Under Policy Management, locate Content Filter and click Edit Policy on the far right.

Step 4. On the policy editing page, enter the keywords you want to block. You can also configure the Action (Block or Alert) and Severity level. Click Save to apply the policy.

Yes. The system supports configuring Exceptions for specific entities:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Security section in the left-hand navigation menu, click Firewall.

Step 3. Locate the target policy in the policy list and click Exceptions on the right.

Step 4. An exception list window will appear. You can add specific users, teams, or agents to the allowlist so that they are exempt from the selected security policy.

Go to Moderation Record to review the detailed blocking history:

Steps:

Step 1. Go to the home page and click Organization Management to open the Organization Management page.

Step 2. Under the Security section in the left-hand navigation menu, click Firewall.

Step 3. Switch to the Moderation Record tab at the top of the page.

Step 4. In this list, you can review complete details of each violation, including the time of occurrence, source, user who triggered the policy, policy name, action taken (such as Alert or Mask), and the corresponding warning message.

This status helps security administrators track the progress of incident handling. After you or your security team reviews and handles a policy violation in Moderation Record, you can click the circle on the far left of the event to mark it as Resolved, making it easier to distinguish alerts that still require attention.

Yes. The Firewall page provides a real-time security statistics dashboard at the top of the page. Whether you are viewing the Policy Management or Moderation Record tab, you can directly review the following information:

Violation Overview: Displays the total number of violations detected over the past several days.
Top Users by Violations: Ranks the employees who triggered the most violations over the past several days, helping the organization provide targeted security awareness training.
Top Violation Types: Displays the policy types triggered most frequently over the past several days, helping you accurately identify the organization's current security weaknesses.

Automation: Automation is designed to standardize repetitive AI workflows. Through this feature, users can automatically execute a series of AI Agents in a customized manner. For enterprises, all automation execution histories are secured within the isolated environment of the Workspace, and can be audited and controlled through organizational management features (recorded in the Automation Log within Org Administration) to balance AI processing efficiency and enterprise data security.

Task: A "repeatable" multi-step operational unit within the automation module. Each task can connect agents, tools, and multiple steps to complete a specific workflow, and features a dedicated Sandbox environment. After creating a task, users can choose to Run and Publish to Automation. Its complete execution status, output results, and error messages are saved in the Run History panel chronologically, facilitating future tracking and re-execution.

You don't necessarily need to build one from scratch. When configuring a task, you can obtain and configure an AI Agent based on your needs through the following methods:

Use system-default Agents and LLMs:
If you only need to execute common basic workflows, you can directly configure standard AI Agents or Large Language Models (LLMs) via Select an agent in the task settings.

Install from the Inventory:
You can first go to the Inventory to browse and install Agents or advanced LLM modules suitable for various specific scenarios into your workspace. Once installed, these resources will be integrated into your environment, allowing you to select them directly when creating automation tasks.

Use custom-built exclusive Agents:
If you have highly customized business needs, you can go to the Agent Studio on the left menu to customize an AI Agent. Once created, you can use this AI Agent to create automation tasks within the same workspace, precisely invoking this exclusive agent tailored for your enterprise.

You can click Select an agent under New Task in Automation to view the capabilities of the agents.

Path: Sidebar → Automation → New Task → Select an agent

View the "Core Capabilities" of Agents and LLMs: The list shows the capabilities and feature tags available for each Agent or LLM. You can check if they possess the following tags based on the core requirements of your task:

Web Search: Suitable for tasks that require real-time internet access to fetch the latest information and perform data verification.
Image Generation: Suitable for workflows that require outputting visual charts or image creation.
Speech-to-Text: Suitable for scenarios involving audio file transcription or voice semantic analysis.
Connector: Suitable for complex tasks that require data linkage with external tools, APIs, or internal enterprise systems.

You can also use the Tags button for one-click filtering. If you know a task requires specific technology, you can directly click the quick filter buttons like Connector or Web Search at the top of the window, and the system will automatically filter out all eligible Agents for you to choose from.

You can click Automation on the left sidebar to create a task and follow these steps:

Step 1. Click Automation on the left sidebar to enter the automation page.

Step 2. Click the Create your first task button on the automation page to go to the new task page. You can customize the automation task name in the New Task title bar at the top of the screen.

Step 3. Click Select an agent. Set up an operating AI Agent or Large Language Model (LLM) for your automation task.

Step 4. Add a Step. You can customize the name of each step based on the task's nature, and use the drag button in the upper right corner to adjust the execution order of each step; click the Duplicate and Delete buttons to copy or remove a step.

Step 5. Enter commands in the Prompt.

Step 6. Add Tools and Connectors. Depending on your task's nature, you can enable tools and connectors for the AI Agent to expand the execution capabilities of that step according to task requirements (available expansion options depend on the native capabilities of the Agent you selected in Step 4).

Step 7. Click Run. Your configured automation task will be executed in the Sandbox on the right. Once the operation is complete, you can view the output results.

Step 8. You can click Rerun and Delete to obtain different output results.

Step 9. Click Publish to Automation to publish your automation task in the current workspace. You and other users with permissions for this workspace can reuse the published automation task in the future.

Before officially executing Publish to Automation, the system provides a secure Sandbox environment for simulation validation:

Step 1. You can review the output content. If it does not meet expectations, you can adjust the prompt, and the system will automatically create a new draft for you.

Step 2. If it still falls short of expectations, you can click Rerun in the Sandbox.

Step 3. You can click the Delete button to remove previously executed drafts.

After verifying the task workflow in the Sandbox environment:

Step 1. Click the Publish button in the upper right corner of the Sandbox page.

Step 2. You can then view the published automation task on the Automation page. You and other users with permissions for this workspace can reuse the published automation task in the future.

This completely depends on the Trigger mechanism you selected during the step configuration:

Schedule: If set to a scheduled timer, the system will automatically start the task in the background strictly according to your specified time cycle (e.g., hourly, daily, or weekly) after publication, requiring no manual intervention.
Manual Trigger: If you choose manual mode, the task will not execute automatically after publication. When you need it to run, you must go to the task list and click the "Play" button (orange icon) on the task card for the task to officially start.

Please go to Automation on the navigation bar or Automation Log in Org Administration:

Automation page: Quickly view and track the execution details and step results of a specific single task.

Step 1. Click to enter the Automation page on the left sidebar.

Step 2. Check the Run History on the page. Click the task card you want to view, and a single record will expand downwards to show detailed execution steps, time spent, as well as the execution summary and specific results of each step.

Automation Log: Global consolidated view, requires Org Administration privileges.

Step 1. Go to Org Administration on the left sidebar.

Step 2. Enter the left menu of Org Administration and click Automation Log. You can view the complete automation execution records within the organization.

Step 3. If you need to view detailed information for a single task, you can click View on the far right of the list.

Please go to Automation on the navigation bar or Automation Log in Org Administration:

Automation page: Quickly view and track the execution details and step results of a specific single task.

Automation Log: Global consolidated view, requires Org Administration privileges.

Step 1. Go to Org Administration on the left sidebar.

Step 2. Enter the left menu of Org Administration and click Automation Log. You can view the complete automation execution records within the organization.

Step 3. If you need to view detailed information for a single task, you can click View on the far right of the list.

Yes. You have full control over the automation tasks you create:

Start/Pause: In the task list, directly click the Start button (play icon) on the task card to quickly toggle the task status.
Modify: Click Edit in the three-dot menu to enter the configuration. For published tasks, it is recommended to re-test in the Sandbox after modification to ensure safety.
Delete: You can delete tasks that are in a draft or paused state. Deleting a task will synchronously clear the task's historical run logs, but it will not affect or delete the AI Agent itself created in the Agent Studio.

If you cannot find a specific AI Agent in the Select an agent menu, please check the following:

Installation status: Confirm whether the agent is installed in the same workspace. You can click Agent Controller to check if the agent has been installed. If not, please go to the Inventory to reinstall.
Publish status: If it is a self-built agent, please go to the Agent Builder to confirm whether it has been successfully published.
Permission settings: If you still cannot find the agent, the organization or workspace administrator may have disabled access to the agent. Please contact your administrator to grant access.

Connectors are the bridge for AI Agents to interact with external systems (such as databases, APIs, communication software). If it shows no available connectors:

Tool configuration: Please go to Agent Studio, check the agent's settings, and confirm whether the corresponding Skills have been added and enabled.
Enable status: Confirm whether the external tools you need (e.g., enterprise databases, Slack, Gmail, etc.) have been successfully connected and are checked as enabled.
Credential verification: If the connector relies on a specific API Key or enterprise account authorization, please ensure that the relevant authorization credentials are correct, still valid, and have not expired.
Language model compatibility: If the above settings are all correct, but the connector still cannot be used or does not display, it may be because the Large Language Model (LLM) currently equipped with the agent does not support calling external tools. It is recommended that you go into the agent settings, change to another language model with higher support, and try again.

When task execution encounters bottlenecks, errors, or the output content does not meet standards, you can locate the problem and fix it through the following methods:

User troubleshooting steps:

Step 1. Go to the Automation page on the left sidebar, find and click the failed or abnormal task card in the Run History on the right.

Step 2. Expand the detailed record and check the Issues field in the Summary section at the bottom to identify where the error occurred.

Step 3. Click Edit in the task card menu to enter the task configuration screen.

Step 4. You can try modifying the Prompt to make the instructions clearer, rechecking the Connector authorization status, or changing to a more suitable Large Language Model (LLM) to improve the logic and quality of the output.

Step 5. After modification, be sure to click Run in the Sandbox on the right for simulation testing. After confirming the results meet expectations, click Publish to Automation to update the schedule.

Advanced troubleshooting steps (Org Administration privileges required):

If general-level information is insufficient to solve the problem, users with administrator privileges can access complete troubleshooting information from the organizational level:

Step 1. Go to the homepage and click Org Administration to enter the organization administration page.

Step 2. Under the Log section on the left, click Automation Log or other related log units.

Step 3. In this section, you can view deeper, more detailed system error causes and execution histories, helping you precisely clarify the root cause of task failures.

The execution of automation tasks will be counted towards the overall usage of your organization. To ensure effective management and transparency of resources:

Consumption calculation: Each time a scheduled trigger or manual execution of a task occurs, the system will calculate the corresponding quota consumption based on the actual processed message count of the AI Agent invoked by the task.

Before setting up high-frequency schedules, it is recommended to perform a test execution in the Sandbox first. The Sandbox will not be included in the calculated consumption. You can also view the overall resource usage status through the Org Administration panel at any time to ensure uninterrupted automation tasks.

ภาษา

ภาษา

If your network connection is unstable, temporarily interrupted, or your device or network environment encounters issues, message delivery may fail, responses may be interrupted, or the connection to the model may be terminated.

We recommend checking your current network connection.

If network latency is high, message transmission and model response times may be prolonged, which may result in slow responses, longer loading times, or what appears to be a disconnection.

Since each message must first go through data governance checks and encryption protection procedures, response time may also be slightly affected under certain circumstances.

We recommend checking your network quality and, if necessary, switching to another network environment before trying again.

On the AnyInsight.ai platform, every message is first processed through a generative AI firewall for data governance checks and encryption protection, which may affect response time.

In addition, if the task is more complex, the selected large language model is under heavy computational load, or the user’s network connection is unstable, responses may become slower or temporarily unresponsive.

To improve overall response efficiency, we recommend simplifying your input, breaking complex requests into smaller parts, and checking your network connection.

Some large language models (LLMs), Agents, Prompt Templates, and MAIA modes may be restricted depending on your subscription plan and the permissions configured by your organization administrator.

As a result, some users may not be able to access certain features. We recommend first checking the permissions currently assigned to your account.

If you still cannot use the feature, you may take the following actions:

Not all AI agents enable MAIA mode by default.

Users can either allow MAIA to adjust automatically based on task complexity or configure it manually.

Additionally, if you choose to upgrade your plan, you will be able to access the services included in the higher-tier plan immediately after the upgrade.

The price difference between your current plan and the upgraded plan will be charged accordingly to ensure fair use and protect user rights.

If you choose to downgrade your plan, the downgrade will take effect automatically after your current plan period ends.

If you have other billing-related questions, you may use the following support channels:

If you do not yet have an AnyInsight.ai account, please first complete the standard user registration process and subscribe to a plan for the first company.

After that, log in to the platform, create a second organization, and subscribe to a plan for the second company.

If you already have an AnyInsight.ai account, you can simply log in, create a second organization, and complete the subscription for that organization.

(If you do not receive the email, please first check your spam folder. If you still have not received it, please try again after one minute.)

Step 5. After email verification is completed, the system will direct you to the account and company information page. Please fill in the information for the first company.

AnyInsight.ai utilizes an auto-renewal mechanism, so you don't have to worry about service interruptions due to plan expiration.

At the end of each billing cycle, the system will automatically process the payment and renew your current subscription.

You may adjust your subscription status at any time on the Manage Plan page. If you cancel your subscription, the service will remain available until the end of the current subscription period.

Notes

Note [1]: Supported file types include PDF, TXT, CSV, DOCX, XLSX, and HTML. Each file can be up to 10 MB, with a maximum of 20 files.

Note [2]: Up to 5 URLs are supported. URL content is automatically updated to the knowledge base weekly. Updates usually take effect within a few minutes; invalid sources will be skipped.

To learn more about AnyInsight.ai’s agent features and use cases, please refer to the following page:

Link: https://www.anyinsight.ai/pages/agent

AnyInsight.ai currently offers three subscription plans: Pro, Growth, and Enterprise, designed to meet the needs of teams and organizations of different sizes. Note [1]

Users can choose the most suitable plan based on their organization size, collaboration needs, security requirements, and intended use cases.