File size: 3,375 Bytes
e6a90e9 b9ddbc7 e6a90e9 b9ddbc7 e6a90e9 b9ddbc7 e6a90e9 b9ddbc7 e6a90e9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 |
---
license: mit
title: 🧠 Multi-Agent AI Assistant
sdk: docker
emoji: 💻
colorFrom: indigo
colorTo: indigo
pinned: true
---
# 🧠 Multi-Agent AI Assistant
This project is a **voice- and text-based AI assistant** powered by **FastAPI**, **LangGraph**, and **LangChain**. It supports:
- 🤖 Multi-agent system
- 🗣️ Voice input via speech-to-text
- 🌄 Image processing
- 📂 File processing
- 🌐 Web search, GitHub actions, weather, time, and geolocation tools
- 🧠 Memory-based conversation history per user
- 📍 Location-aware capabilities (e.g., timezone, weather)
---
## 🚀 Features
- **FastAPI backend** with `/text`, `/image`, `/voice` and `/file` endpoints
- **LangGraph agents** with memory, tools, and typed state
- **User database** with persistent settings and config
- **Agents**:
- Supervisor agent
- Deep research agent
- Coding agent
- **Supervisor tools**:
- Web search
- Current time
- Weather
- Yahoo finance news
- **Deep research tools**:
- Web search
- Wikipedia
- Yahoo finance news
- **Coder tools**:
- Github toolkit
- Web search (For docs research)
---
## 🛠️ Setup Instructions
### 1. Environment
This project requires no manual environment configuration — all secrets are passed dynamically with each request and stored in the database.
Just run:
```bash
docker build -t multi-agent-assistant .
docker run -p 7860:7860 multi-agent-assistant
```
The server will launch automatically at http://localhost:7860, with all dependencies installed and configured inside the container.
---
## 📦 API Reference
### `/text` – Send Text Prompt
```http
POST /text
Form data:
- state (str): JSON-encoded state dict
```
### `/voice` – Send Audio Prompt
```http
POST /voice
Form data:
- state (str): JSON-encoded state dict
- file (binary): Audio file (WAV, MP3, etc.)
```
### `/image` – Send Image Prompt
```http
POST /image
Form data:
- state (str): JSON-encoded state dict
- file (binary): Image file (JPEG, PNG, etc.)
```
### `/file` – Send File Prompt
```http
POST /file
Form data:
- state (str): JSON-encoded state dict
- file (binary): Image file (PDF, TXT, etc.)
```
---
## 🧩 Agent State Structure
The system operates using a JSON-based shared state object passed between agents.
It follows a `TypedDict` schema and may contain keys such as:
```json
{
message: AnyMessage (message to the agent)
user_id: str (unique user id)
first_name: str
last_name: str
assistant_name: str
latitude: str
longitude: str
location: str (user-readable location)
openweathermap_api_key: str
github_token: str
tavily_api_key: str
groq_api_key: str
clear_history: bool (True/False)
messages: list (Filled automatically from the database)
}
```
This dictionary acts as a single mutable state shared across all agent steps, allowing for data accumulation, tool responses, and message tracking.
---
## 🧠 Built With
- [Groq](https://github.com/groq/groq-python)
- [LangGraph](https://github.com/langchain-ai/langgraph)
- [LangChain](https://github.com/langchain-ai/langchain)
- [FastAPI](https://fastapi.tiangolo.com/)
- [Whisper](https://github.com/openai/whisper)
- [SQLAlchemy + Async](https://docs.sqlalchemy.org/en/20/orm/extensions/asyncio.html)
---
## 📝 License
MIT — Feel free to use, modify, and contribute!
|