Introduction to ChatPilot: Chat Agent Web UI
ChatPilot is an innovative web-based application designed to facilitate seamless interactions through chat agents. It is a comprehensive platform that supports advanced functionalities such as Google search, URL-based conversation, and code execution, making it a versatile tool for diverse applications. ChatPilot is built to integrate with multiple AI models and APIs, enhancing its usability across different domains.
Key Features
Agent Assistant Capabilities:
ChatPilot is developed using the Agentica framework, enabling robust Agent Assistant functionalities. It offers:
-
Tool Integration: ChatPilot allows agents to interface with external tools, including:
- Web Search Tools: Utilizing Google Search API with options like Serper and DuckDuckGo.
- URL Parsing: Replicates the functionality of Kimi Chat for seamless URL-based interactions.
- Python Code Interpreter: Supports code execution in both local and virtual environments such as E2B.
-
AI Model Integration: The platform supports the integration of multiple Language Models (LLMs) through various APIs:
- Options include local open-source models via Ollama Api, cloud-deployed models via litellm Api, and GPT models via OpenAI Api.
-
RAG Support: ChatPilot features RAG (Retrieval-Augmented Generation) for enhanced agent-driven document querying.
Flexible Architecture:
ChatPilot's architecture separates frontend and backend services. The frontend is built using Svelte while the backend leverages FastAPI, ensuring scalability and performance.
Enhanced User Experience:
The platform supports features like voice input/output, image generation, user management, access control, and import/export of chat logs.
Demonstration
To witness ChatPilot in action, users can explore the official demo, showcasing its functionality and user interface.
Getting Started
Running ChatPilot with Docker:
ChatPilot can easily be deployed using Docker, with simple commands to set up environment variables and run the application locally or on a server.
Local Installation:
For users preferring manual installation, the source code is available for cloning from GitHub. After setting up dependencies and configurations, ChatPilot can be executed locally.
Frontend Development:
Developers have two options for the frontend: utilize the pre-compiled build or customize the frontend code using Node.js, followed by a build process with npm.
Integration with Various LLMs
OpenAI API:
By configuring environment variables, ChatPilot supports integration with GPT models using the OpenAI API. It also provides options for integrating Azure's version of OpenAI API with specific configurations.
Ollama API:
For local open-source models, ChatPilot can integrate through Ollama API, facilitating diverse model usage based on user requirements.
litellm API:
This option allows integration with cloud-deployed models, providing flexibility in choosing model sources and configurations.
Contact and Contributions
For feedback, issues, or contributions, users are encouraged to reach out through the GitHub issues page or directly via the provided contact details. Contributions are welcomed, provided they include appropriate unit tests and pass all test checks.
ChatPilot is a sophisticated tool designed for modern conversational needs, leveraging AI and web technologies to offer versatile chat solutions. Its ability to integrate diverse models and APIs makes it a valuable resource for developers and businesses aiming to enhance their interactive capabilities.