⚡ Fastapi Langgraph Chatbot

A clean, modern, and lightweight local chatbot built with FastAPI and LangGraph.

Designed for effortless local execution with zero complex dependencies (no external Docker, PostgreSQL, or Redis required).

---

✨ Features

• 🚀 Instant Local Run: Runs directly on http://localhost:8000.
• 🧠 Multi-Provider LLM Support:
  • Groq: Ultra-fast openai/gpt-oss-120b, llama-3.3-70b-versatile
  • Ollama: 100% free local LLMs running privately on your machine (llama3.2, mistral, deepseek-r1, qwen2.5).
  • OpenAI: gpt-4o, gpt-4o-mini
  • Google Gemini: gemini-1.5-flash, gemini-2.0-flash
  • Anthropic: claude-3-5-sonnet
• 💾 Database Persistence (SQLModel / SQLite / PostgreSQL):
  • Persistent chat history and session management saved to local chatbot.db.
  • Zero-setup SQLite by default, with seamless support for PostgreSQL via DATABASE_URL.
• 📚 Retrieval-Augmented Generation (RAG):
  • Upload .txt, .pdf, and .md documents via API or UI.
  • Asynchronous, background pipeline for intelligent chunking, deduplication, and vector storage (powered by local Qdrant).
  • Context-aware retrieval for specific sessions/users.
• 🔄 Rolling Memory & Summarization:
  • Conversation summarization using LangGraph to keep state context within LLM token limits without losing important facts.
• 🛠️ Built-in Agent Tools (LangGraph):
  • Math Evaluator (calculate)
  • Real-time Date/Time tool (get_current_time)
  • Web Search (search_web) via DuckDuckGo
• 💬 Real-Time Streaming: Server-Sent Events (SSE) for token-by-token instant typing.
• 🎨 Glassmorphic Web UI:
  • Dark / Light mode toggle
  • Code syntax highlighting with copy buttons
  • Chat history session manager synced with database
  • Markdown rendering (tables, bold, lists, code blocks)
  • Export chat history as Markdown

---

🚀 Quick Start

1. Clone the Repository

git clone https://github.com/ojask923/Chatbot.git
cd Chatbot

---

2. Run the Application

Option A: Double-Click (Windows - Easiest)

Simply double click run.bat. It will automatically create a .venv, install all dependencies, and open http://localhost:8000 in your browser.

Option B: Command Line

# 1. (Optional) Create and activate virtual environment
python -m venv .venv
.\.venv\Scripts\Activate.ps1

# 2. Install requirements
pip install -r requirements.txt

# 3. Start the server
python start.py

Open http://localhost:8000 in your browser!

---

⚙️ Configuration & API Keys

Copy .env.example to .env:

cp .env.example .env

Edit .env to configure your preferred provider:

# Default provider: groq | ollama | openai | gemini | anthropic
DEFAULT_PROVIDER=groq
DEFAULT_MODEL=openai/gpt-oss-120b

# Set your API keys (if using cloud providers)
GROQ_API_KEY=your_groq_key_here
OPENAI_API_KEY=your_openai_key_here
GEMINI_API_KEY=your_gemini_key_here
ANTHROPIC_API_KEY=your_anthropic_key_here

# Database URL (defaults to local SQLite database chatbot.db, or postgresql://user:pass@host/db)
DATABASE_URL=sqlite:///./chatbot.db

---

📁 Project Structure

├── app/
│   ├── agent/
│   │   ├── graph.py            # LangGraph state machine & multi-provider factory
│   │   └── tools.py            # Math, time, and web search tools
│   ├── api/
│   │   ├── routes.py           # FastAPI chat, stream, history & session endpoints
│   │   └── rag_routes.py       # RAG document ingestion & retrieval admin endpoints
│   ├── models/                 # Database Models (SQLModel)
│   │   ├── session.py          # ChatSession table
│   │   └── message.py          # ChatMessage table
│   ├── services/               # Services Layer
│   │   ├── database.py         # Database connection & CRUD operations
│   │   ├── memory.py           # Mem0 user preferences & long-term memory
│   │   ├── rag_service.py      # Document parsing, chunking, and vector embedding
│   │   ├── summarizer.py       # Rolling conversation compactor for AgentState
│   │   ├── context_engine.py   # Advanced context retrieval logic
│   │   ├── query_rewriter.py   # Query rewriting for optimal vector search
│   │   └── qdrant_shared.py    # Qdrant client connection handling
│   ├── static/                 # Modern Web Chat UI
│   │   ├── index.html          # HTML Layout
│   │   ├── style.css           # Glassmorphism styling & animations
│   │   └── app.js              # SSE streaming & database-synced session manager
│   └── config.py               # Pydantic environment configuration
├── chatbot.db                  # Local SQLite database file (auto-generated)
├── main.py                     # FastAPI app initialization & database lifespan
├── start.py                    # Launcher script with auto browser popup
├── run.bat                     # Windows 1-click batch launcher
├── requirements.txt            # Dependencies
└── .env                        # Environment config