AI & LLM Systems
End-to-end RAG pipelines, Text-to-SQL engines with semantic layers, intent classifiers, hallucination mitigation stacks, and real-time SSE token streaming. Powered by NVIDIA NIM, LlamaIndex, and FastAPI.
Full Stack Engineer at Techinterio, building RAG pipelines, Text-to-SQL engines, and enterprise HR SaaS. Based in Noida, India.
End-to-end RAG pipelines, Text-to-SQL engines with semantic layers, intent classifiers, hallucination mitigation stacks, and real-time SSE token streaming. Powered by NVIDIA NIM, LlamaIndex, and FastAPI.
Production-grade monorepo platforms with end-to-end type safety, real-time WebSocket communication, multi-role RBAC, CI/CD pipelines, and cross-platform mobile delivery from a single codebase.
Production HR SaaS serving 400+ employees. 93 end-to-end type-safe tRPC procedures, 429 UI components, Casbin RBAC with 4 role tiers, full audit logging, and real-time bidirectional chat via Socket.io.
6-step Text-to-SQL pipeline with governed semantic layer. 3-tier hybrid NLU classifier (~40 intents), multi-backend PDF OCR, real-time SSE streaming, and multi-layer hallucination mitigation.
Open-source AI CLI coding assistant with Bun and TypeScript supporting code generation, refactoring, and prompt management. Companion Next.js web interface with real-time prompt editing.
Full-stack rental platform with 40+ listings, calendar-based booking, Leaflet/Google Maps integration, Recharts analytics, PostHog tracking, and JWT auth backed by MongoDB.
Headless e-commerce platform with 15+ product SKUs, cart, and checkout. Next.js frontend connected to Shopify headless CMS via REST API with full TypeScript coverage and Zod validation.
Social media platform for fundraising with post feeds, investor matching, community engagement, and real-time notifications. Page load times under 2 seconds with full TypeScript coverage.
Open to full-time roles, freelance projects, and interesting collaborations in AI systems and full-stack engineering.