yash.dev
I build AI-native products end to end — multi-tenant RAG platforms, agentic tool-calling concierges, and production web apps. Next.js + TypeScript frontends, Node.js APIs, PostgreSQL/pgvector & MongoDB underneath.
CS undergrad and Full-Stack AI Engineer. I ship AI-native products end to end — RAG pipelines with pgvector and semantic search, agentic tool-calling loops on Groq LLaMA, and the Next.js/Node.js apps that wrap them.
TypeScript by default, prompts engineered like APIs, reliability treated as a feature. Previously interned as a Frontend Developer at Ai-Learn, Bengaluru.
Multi-tenant RAG platform with workspace-scoped document isolation, RBAC, and a tool-calling chat system delivering cited, token-streamed answers over your documents.
Hotel booking platform with a real-time agentic concierge — a multi-turn tool-calling loop on LLaMA 3.3-70B that searches inventory, checks availability, and books rooms.
Full-stack AI travel platform. Generate personalized day-by-day itineraries with hotels, restaurants & budgets via natural-language prompts.
Comprehensive collection of projects, examples & notes on AI app development with Next.js 15 and Vercel AI SDK v6 — LLM integration, streaming, embeddings, vector search, multimodal AI, and production-ready workflows.
Scalable REST API for video & tweet management — 25+ endpoints, MVC architecture, JWT auth with access/refresh token rotation.
Whether you have an AI product in mind, a role to fill, or just want to talk RAG pipelines and agentic flows — my inbox is always open.