DevAssistAI

DevAssistAI is a GenAI-powered developer assistant that supports code generation, debugging, code explanation, and document-based Q&A using RAG.

Tech Stack

• Backend: FastAPI, Python
• LLM: Gemini API
• RAG: ChromaDB, LangChain-ready architecture
• Database: SQLite for query and latency logging
• Frontend: Streamlit
• Deployment-ready: AWS-compatible backend structure

Features

• Generate production-style code from natural language prompts
• Debug code snippets and error logs
• Explain unfamiliar code in simple language
• Upload documents and ask grounded questions using RAG
• Store query logs and latency metrics in SQLite
• Clean Streamlit UI for quick demos

Project Structure

DevAssistAI/
├── backend/
│   ├── main.py
│   ├── llm.py
│   ├── rag.py
│   ├── db.py
│   ├── config.py
│   └── schemas.py
├── frontend/
│   └── app.py
├── data/
├── tests/
├── .env.example
├── .gitignore
├── requirements.txt
└── README.md

Setup

1. Clone the repo

git clone https://github.com/YOUR_USERNAME/DevAssistAI.git
cd DevAssistAI

2. Create virtual environment

python -m venv .venv
source .venv/bin/activate

For Windows:

.venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Add environment variables

cp .env.example .env

Then edit .env:

GEMINI_API_KEY=your_gemini_api_key_here
DATABASE_URL=sqlite:///./devassist.db
CHROMA_DIR=./chromadb_store
MODEL_NAME=gemini-1.5-flash
BACKEND_URL=http://localhost:8000

Run the App

Start FastAPI backend

uvicorn backend.main:app --reload

Backend will run at:

http://localhost:8000

API docs:

http://localhost:8000/docs

Start Streamlit frontend

Open a new terminal:

source .venv/bin/activate
streamlit run frontend/app.py

Frontend will run at:

http://localhost:8501

API Endpoints

Endpoint	Method	Purpose
/generate	POST	Generate code from a prompt
/debug	POST	Debug code or error logs
/explain	POST	Explain code
/rag/upload	POST	Upload and ingest documents
/rag/query	POST	Ask questions from uploaded documents
/logs	GET	View query logs and latency