Chatbot Assistant with Gemini + gRPC

This project is a simple chatbot backend powered by Google's Gemini API and uses gRPC for client-server communication. It maintains context per session and provides interactive conversational experiences.

Architecture

Features

• gRPC-based communication
• Google Gemini API integration
• Session-based context management
• Clean modular design (client, server, context manager, inference engine)

📂 Project Structure

assistant_backend/
│
├── assets/
│   └── image.png                # (Optional) Image asset
│
├── protos/
│   └── (Protobuf files for gRPC definitions)
│
├── assistant_pb2.py            # Generated from .proto
├── assistant_pb2_grpc.py       # Generated from .proto
├── client.py                   # gRPC client
├── server.py                   # gRPC server with Gemini integration
├── context_manager.py          # Manages chat context per session
├── inference_engine.py         # Contains the logic to generate a reply
├── requirements.txt            # Python dependencies
└── .env                        # Gemini API Key

1. Install Dependencies

pip install -r requirements.txt

2. Set Up Environment

Create a .env file with your Gemini API key:

GEMINI_API=your_api_key_here

3. Run the Server

python server.py

4. Run the Client

python client.py

Example Usage

Type 'exit' to stop chatting.
You: Hello
AI: I received: Hello. Previous context: 0 messages.

Notes

• Ensure your .proto files are compiled to assistant_pb2.py and assistant_pb2_grpc.py
• Python version 3.10 or higher is recommended

Built With

• Python
• gRPC
• Google Gemini API
• dotenv

© 2025 Om Jamnekar. All rights reserved.