Chat Interface

Tyk AI Studio’s Chat Interface provides a secure and interactive environment for users to engage with Large Language Models (LLMs), leveraging integrated tools and data sources. It serves as the primary front-end for conversational AI interactions within the platform.

Purpose

The main goals of the Chat Interface are:

• User-Friendly Interaction: Offer an intuitive web-based chat experience for users of all technical levels.
• Unified Access: Provide a single point of access to various configured LLMs, Tools, and Data Sources.
• Context Management: Maintain conversation history and manage context, including system prompts and retrieved data (RAG).
• Secure & Governed: Enforce access controls based on teams and apply configured Filters.

Key Features

• Chat Sessions: Each conversation happens within a session, preserving history and context.
• Streaming Responses: LLM responses are streamed back to the user for a more interactive feel.
• Tool Integration: Seamlessly uses configured Tools when the LLM determines they are necessary to fulfill a user’s request. The available tools depend on the Chat Experience configuration and the user’s group permissions.
• Data Source (RAG) Integration: Can automatically query configured Data Sources to retrieve relevant information (Retrieval-Augmented Generation) to enhance LLM responses. The available data sources depend on the Chat Experience configuration and the user’s group permissions.
• System Prompts: Administrators can define specific system prompts for different Chat Experiences to guide the LLM’s persona, tone, and behavior.
• History: Users can view their past chat sessions.
• File Upload (Context): Users might be able to upload files directly within a chat to provide temporary context for the LLM (depending on configuration).
• Access Control: Users only see and can interact with Chat Experiences assigned to their Teams.

Using the Chat Interface

Users access the Chat Interface through the Tyk AI Studio web UI.

1. Select Chat Experience: Users choose from a list of available Chat Experiences (pre-configured chat environments) they have access to.

2. Interact: Users type their prompts or questions.

3. Receive Responses: The LLM processes the request, potentially using tools or data sources behind the scenes, and streams the response back.

   

Configuration (Admin)

Administrators configure the available “Chat Experiences” (formerly known as Chat Rooms) via the UI or API. Configuration involves:

• Naming: Giving the Chat Experience a descriptive name.

• Assigning LLM: Linking to a specific LLM Configuration.

• Enabling Tools: Selecting which Tool Catalogues are available.

• Enabling Data Sources: Selecting which Data Source Catalogues are available.

• Setting System Prompt: Defining the guiding prompt for the LLM.

• Applying Filters: Associating specific Filters for governance.

• Assigning Groups: Determining which Teams can access this Chat Experience.

• Enabling/Disabling Features: Toggling features like file uploads or direct tool usage.

  

API Access

Beyond the UI, Tyk AI Studio provides APIs (/api/v1/chat/...) for programmatic interaction with the chat system, allowing developers to build custom applications or integrations that leverage the configured Chat Experiences.

This comprehensive system provides a powerful yet controlled way for users to interact with AI capabilities managed by Tyk AI Studio.