Seeyon Chat User Manual

I. Introduction

1. What is Seeyon Chat?

Seeyon Chat is an innovative platform that seamlessly integrates cutting-edge AI technology into your daily workflow. It empowers you to create intelligent chatbots, streamline collaborative processes, and leverage the power of large language models (LLMs) to enhance productivity and efficiency.

2. Key Features and Benefits

3. Target Audience

Seeyon Chat is designed for a wide range of users, including:

II. Getting Started

1. Account Registration and Login

To access Seeyon Chat, you need to register for an account.

2. User Interface Overview

Once logged in, you'll be greeted by the Seeyon Chat dashboard. The interface is designed for ease of use and intuitive navigation.

3. Navigating the Dashboard

You can easily switch between tabs by clicking on the corresponding tab in the navigation bar.

III. Chatbots

1. Understanding Chatbots

Chatbots are AI-powered conversational agents that can interact with users in a natural language format. They can be programmed to provide information, answer questions, automate tasks, and engage in conversations. Seeyon Chat leverages the power of LLMs to create intelligent and versatile chatbots.

2. Creating a Chatbot

a. Defining Chatbot Details

To create a new chatbot, navigate to the "Workshop" tab and click on the "Create Chatbot" button. You'll be presented with a form to define the chatbot's details:

b. Using the AI Assistant Description Generator

If you need help crafting a compelling chatbot description, Seeyon Chat provides an AI-powered description generator.

c. Adding Files and APIs

You can enhance your chatbot's knowledge and capabilities by adding relevant files and integrating external APIs.

d. Enabling Tools

Seeyon Chat allows you to enable specific tools for your chatbot, expanding its functionalities beyond simple conversation. These tools can be accessed and utilized by the LLM based on the context of the conversation.

e. Setting Up Forms

Forms enable you to collect structured data from users and leverage it to personalize the chatbot's responses upon start of a new chat.

f. Managing Members deprecated

For collaborative projects, you can add members to your chatbot and assign them specific roles.

g. Publishing and Unpublishing

Once you're satisfied with your chatbot, you can publish it to make it available to other users.

Your submitted chatbots will be reviewed by an administrator before being published.

3. Cloning Chatbots

You can clone existing chatbots to create new ones based on their settings and configurations.

4. Interacting with Chatbots

a. Starting a Chat

To interact with a chatbot, click on its name or avatar in the "Home" or "Market" tab. This will open a new chat window.

b. Sending Messages and Files

c. Using Slash Commands

Slash commands are shortcuts that allow you to quickly access chatbot features or perform actions.

d. Regenerating Responses

If you're not satisfied with a chatbot's response, you can regenerate it by clicking the "Regenerate" button. This will prompt the chatbot to generate a new response based on the same input.

e. Special Element Renderers

Seeyon Chat features special element renderers that allow chatbots to generate and display interactive elements directly within the chat interface. These elements can be created using React, Vue, or the v5PortalWidget framework.

When the chatbot generates a response containing these special elements, you'll see a collapsible section titled "Tool call completed." Expanding this section will reveal either the live preview of the component/widget or the raw JSON code, depending on a toggle switch.

Consult the Seeyon Chat Developer Document for more details on how to create and use these special element renderers.

5. Viewing Chatbot Details

You can view the details of a chatbot by clicking on its name in the "Workshop" or "Market" tab. The detail page displays information about the chatbot, including its description, model, prompt, and members.

6. Finding Chatbots in the Market

a. Browsing and Filtering

The "Market" tab displays a list of published chatbots. You can browse through the list or use filters to narrow down your search.

b. Starring Chatbots

You can star chatbots that you find useful or interesting. Starred chatbots will be displayed prominently on your "Home" tab for easy access.

IV. Chats

1. Managing Chats

a. Creating a New Chat

You can start a new chat with any chatbot from the "Home" or "Market" tab. Simply click on the chatbot's name or avatar, and a new chat window will open. You can also initiate a new chat without selecting a specific chatbot, which allows you to use slash commands to choose a chatbot within the chat window.

The Home tab also provides a chat box at the bottom of the page, allowing you to start a new chat with the default LLM.

b. Renaming Chat Titles

You can rename a chat to make it easier to identify in your chat history.

The chat title is generated based on the chatbot's description and the last message in the chat.

c. Deleting Chats

You can delete chats from your chat history if you no longer need them.

2. Chat History and Navigation

Seeyon Chat provides convenient ways to manage and navigate your chat history.

3. Exporting Chat Content

You can export the content of a chat in various types for sharing, archiving, or further analysis.

4. Forking Conversations

The "Fork" feature allows you to create a new chat from an existing conversation, starting from a specific point in the chat history. This is useful for exploring alternative paths or focusing on a particular aspect of the conversation.

V. Advanced Features

1. API Integration

Seeyon Chat allows you to integrate external APIs into your chatbots, expanding their functionality and enabling them to access real-time information and interact with external systems.

a. Enabling API Access

b. Adding OpenAPI Specifications

c. Referencing APIs in Chatbot Prompts

To make your chatbot aware of the available APIs, you can reference them in its prompt.

{{#API}}
Available APIs:
{{API}}
{{/API}}

This will display a list of available APIs within the chatbot's prompt, making it aware of the potential functions it can call.

Since inclusion of the any existing API information in the prompt is the default behavior, you can omit the {{#API}}...{{/API}} tags if no additional API information is needed.

d. Triggering API Calls

API calls are triggered based on the user's input and the chatbot's understanding of the available APIs. The chatbot will automatically generate the necessary code using the httpRequest function to interact with the API.

2. Prompt Template Customization

Seeyon Chat uses Mustache, a logic-less template engine, to create dynamic and personalized chatbot prompts.

a. Understanding Mustache Syntax

b. Using Variables in Prompts

Seeyon Chat provides a variety of variables that can be used within prompts:

Variable Description
chatContext Contextual details like form data, files, and code blocks.
formData Data submitted through forms.
form Serialized form data.
chatbot Chatbot properties like name or state.
user User information like name or preferences.
memories Serialized memory data. (Deprecated)
empty Boolean flag indicating an empty conversation.
today Current date in Chinese format.
API Descriptions or lists of available APIs.
currentUserRole Current user's role. (Deprecated)
followingToolResult Boolean flag indicating if the last message was a tool result.

c. Best Practices for Prompt Design

For more details on how to design prompt templates, see the Prompt Template Customization document.

3. Form Creation and Usage

Forms allow you to gather structured data from users, making your chatbot interactions more efficient and personalized.

a. Defining Form Fields

b. Using the Write for Me Feature

If you're unfamiliar with JSON Schema, you can use the "Write for Me" feature.

c. Referencing Form Data in Chatbot Responses

Use Mustache syntax to reference form data in your chatbot's prompt. For example, {{formData.name}} will render the user's submitted name.

4. Collaborator Management deprecated

You can collaborate with others on chatbot development by adding collaborators to your project.

a. Adding and Removing Collaborators

VI. Administration

Seeyon Chat offers robust administration tools to manage users, groups, and access system-wide metrics and analytics.

1. User Management

a. Viewing User Statistics

Navigate to the "Admin" tab and select "Users."

alt text

You can view a list of all registered users, along with their roles, activity status, and other relevant information.

b. Searching for Users

Use the search bar to quickly find specific users by name or email address.

c. Editing User Roles and Activity Status

As an administrator, you can modify user roles and activate or deactivate accounts.

Select the desired role from the dropdown menu next to the user's name.

Toggle the "Is Active" switch to activate or deactivate the account.

2. Group Management

a. Creating and Editing Groups

Navigate to the "Admin" tab and select "Groups."

alt text

Click the "Create Group" button to create a new group.

To edit an existing group, click on its name in the group list.

You can modify the group's name, contact information, and designate administrator users.

b. Managing Group Invitation Codes

Each group has a unique invitation code that can be used to invite new members.

To reset the invitation code, click the "Reset Invitation Code" button.

c. Adding Credits to Groups

Groups have a credit balance that determines their usage limits.

Administrators can add credits to a group by clicking the "Add Credits" button and specifying the amount.

3. System Metrics and Analytics

Seeyon Chat provides comprehensive analytics to track system usage, chatbot popularity, and user engagement.

a. Viewing User Registration Trends

The "Registered By Date" chart displays the number of users registered each day, segmented by active and inactive users.

alt text

b. Analyzing User Message Frequency

The "User Messages by Date" chart shows the total number of messages sent by users over time.

alt text

c. Identifying Top Users

The "Top Users" chart ranks users based on the number of messages they have sent.

alt text

d. Tracking Chatbot Usage and Popularity

The "Most Starred Chatbots" and "Most Popular Chatbots" charts showcase the chatbots that are most frequently starred and used by users, respectively.

alt text

alt text

VII. Settings and Preferences**

1. Language Selection

2. API Key Management

3. Feedback and Support

VIII. Appendix**

1. Built-in Models and Agents

Seeyon Chat offers a wide range of built-in models and agents, empowering you to build diverse and powerful chatbots.

a. Large Language Models (LLMs)

These models are the core of Seeyon Chat's conversational AI capabilities, enabling chatbots to understand and generate human-like text.

b. Multimodal Models

These models go beyond text, allowing chatbots to process and generate both text and images.

c. Text-to-Image Agents

These agents are specifically designed for generating images from text descriptions.

Note: The availability of specific models and agents may vary depending on your user role and the deployment configuration of Seeyon Chat.

1. Glossary of Terms

Term Definition
AI (Artificial Intelligence) The ability of a computer or a robot controlled by a computer to do tasks that are usually done by humans because they require human intelligence and discernment.
API (Application Programming Interface) A set of definitions and protocols that allows one application to access the features or data of another application.
Chatbot An AI-powered conversational agent that can interact with users in a natural language format.
Collaboration The action of working with someone to produce or create something.
Fork The creation of a new chat from an existing conversation, starting from a specific point in the chat history.
JSON Schema A vocabulary that allows you to annotate and validate JSON documents.
LLM (Large Language Model) A deep learning algorithm that can recognize, summarize, translate, predict, and generate text and other content based on knowledge gained from massive datasets.
Mustache A logic-less template syntax used for generating dynamic content.
OpenAPI Specification A standard for defining and documenting RESTful APIs.
Prompt An input given to an LLM to guide its behavior and generate a response.
Role A set of permissions and access levels assigned to a user within a chatbot or group.
Slash Command A shortcut used in the chat input box to trigger specific actions or features.
Stream To transmit or receive data over a network as a continuous flow.
Token A unit of text used to measure the length of prompts and responses in LLMs.

2. Frequently Asked Questions

3. Contact Information

▶︎
all
running...