AI and local knowledge
I’m interested in AI that is useful, inspectable, and respectful of the data it works with. Documind explores private local RAG workflows; DrugForgeRL explores LLM agents in scientific environments.
Hello! I’m Vedant Vijay.
I build practical software across AI, web, and mobile.
I’m a full-stack developer who enjoys turning ambitious ideas into usable products. My public work spans local AI tools, retrieval-augmented generation, AI-driven drug discovery, safety applications, sustainability apps, and algorithm practice.
I care about the details that make software feel trustworthy: clear interfaces, grounded outputs, sensible data flows, and enough engineering underneath the surface for the product to keep working.
What I keep coming back to
Product engineering
I like owning the full path from a rough idea to a working interface: modeling the problem, building the API, shaping the interaction, and making the details feel intentional.
Learning by shipping
My GitHub is a record of experiments, algorithms, interfaces, and products built to answer real questions through code.
Stack
- TypeScript
- JavaScript
- Python
- Swift
- Java
- C++
- React
- Node.js
- Vite
- LangChain
- FAISS
- Ollama
- Streamlit
- Docker
- Firebase
- MongoDB
- Git