Building AI-native products with TypeScript, Python, Next.js, PostgreSQL, and LLMs.
I build AI-native software where LLMs and embeddings are core parts of the system.
- π‘οΈ Building Siftguard β AI-powered GitHub App for issue & PR triage
- πΊοΈ Building Liner β Visual roadmap & habit tracker
- π± Open Source contributor to LangGraph
- π» TypeScript β’ Python β’ React β’ Next.js β’ PostgreSQL
- π Currently learning ML fundamentals, retrieval systems, and evaluation
LLM Integration
Prompt Engineering
Embeddings
RAG
Vector Search
pgvector
Semantic Search
AI Evaluation
Anthropic Claude
shadcn/ui β’ React Flow β’ Framer Motion β’ Zustand
REST APIs
GitHub Apps
GitHub Actions
Row-Level Security
AES-256-GCM
AI-powered GitHub App for OSS maintainers.
Highlights
- Embedding-based duplicate detection
- AI quality scoring
- Cost-aware LLM pipeline
- GitHub Actions integration
- Secure encrypted secrets
Stack
TypeScript
Probot
Anthropic Claude
PostgreSQL
pgvector
π https://github.com/Navneet-Scaler/Siftguard
Interactive roadmap and habit tracking application.
Highlights
- Zoomable learning roadmap
- GitHub-style heatmap
- Outline import
- Supabase RLS
- JSON backup & restore
Stack
Next.js
React
TypeScript
Supabase
π https://liner-xi.vercel.app
π https://github.com/Navneet-Scaler/Liner
Python scraper for Amazon & Flipkart product information.
Python
BeautifulSoup
π https://github.com/Navneet-Scaler/Python-Based-Web-Scrapper
- LangGraph contributions
- Bug reports
- Fixes
- Documentation improvements
Built with β and a healthy respect for pgvector.


