sankalp1999/code_qa
RAG on codebases using treesitter and LanceDB
Extracts code structure via tree-sitter AST parsing to enable language-agnostic indexing across Python, Rust, JavaScript, and Java, then stores embeddings in LanceDB for vector similarity search. Implements a two-stage retrieval pipeline with optional ColBERT reranking to surface contextually relevant code chunks, powered by OpenAI/Jina embeddings and LLM-based chat responses. Built on Flask with Redis caching, supporting interactive natural language code exploration through a web UI.
281 stars. No commits in the last 6 months.
Stars
281
Forks
42
Language
Python
License
MIT
Category
Last pushed
Nov 17, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/sankalp1999/code_qa"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
brevia-ai/brevia
Extensible API and framework to build your Retrieval Augmented Generation (RAG) and Information...
swirlai/swirl-search
AI Search & RAG Without Moving Your Data. Get instant answers from your company's knowledge...
wzdavid/ThinkRAG
A LLM RAG system runs on your laptop. 大模型检索增强生成系统，可以轻松部署在笔记本电脑上，实现本地知识库智能问答。企业级SaaS版本请访问：
thinkany-ai/rag-search
RAG Search API
Piazza-tech/Piazza-Updater
Piazza-Updater automates updates to a Weaviate database with real-time vectorial data. By...