AI Engineer building Generative AI retrieval and agent systems that hold up in production.
Pune, India
Multi-provider LLM orchestration for Vsynergize AI. 2025, Production.
Production-grade Generative AI agents in Python for enterprise workflows — prompt engineering and multi-provider LLM orchestration tuned for accuracy, tone and reliability.
Python, FastAPI, AWS Bedrock
Browser-based voice calling with real-time AI pipelines. 2025, Production.
Built core AI features for a browser-based calling platform — outbound and inbound voice agents (real-time STT → LLM → TTS) supporting any number of configurable campaigns, with multi-agent handoff that routes complex sub-tasks to specialized agents mid-call. Plus a live AI call copilot that coaches human agents with RAG-grounded, citation-backed suggestions, call transcription and sentiment analysis, and AI-scored LinkedIn lead sourcing. 90K+ lines of code, 400+ automated tests.
LiveKit, Plivo, AWS Bedrock, RAG
Backend lead on a production learning management system. 2025, Production.
Led backend development of a multi-role learning platform serving Admins, Faculty, Students and Parents — 500+ REST endpoints across 90K+ lines of async Python, with dual-gateway subscription billing (Razorpay and Stripe, webhook-verified) and 2,100+ automated tests, deployed serverless on AWS Lambda.
FastAPI, PostgreSQL, Razorpay/Stripe
Structured LLM summaries via AWS Bedrock. 2025, Production.
An agent that retrieves insurer data and generates structured, LLM-driven summaries to automate customer interactions.
AWS Bedrock, RAG, Lambda
Generative AI for agricultural pollination monitoring. 2025, Research — Annam AI, IIT Ropar.
Applied Generative AI and ML (Scikit-learn, TensorFlow) to help farmers improve crop yield through pollination monitoring.
Python, Scikit-learn, TensorFlow
Intent-first social discovery app for Pune. 2026, Solo project — pre-launch beta.
Solo-built end-to-end: a React Native mobile app on a Supabase backend (Postgres, RLS, Realtime) with Cloudflare R2 media storage. Security-first by design — every table default-deny RLS, 28 automated policy tests in CI before every migration. 11K+ lines of TypeScript across 10 feature areas, 17 migrations, 2,800+ lines of SQL.
React Native, Supabase, PostgreSQL RLS
Turns bank & UPI SMS into a spending ledger, fully on-device. 2026, Solo project — released.
An Android app that reads bank and UPI SMS and turns them into categorized Transactions, with monthly trends, Merchant and Category breakdowns and CSV/JSON export. No server, no account: data lives in on-device SQLite, and release builds strip every permission except READ_SMS, including network access. A native Kotlin SMS module under an Expo app, pure-TypeScript domain logic with 200+ automated tests, and APKs built and published by GitHub Actions on every tag.
React Native, Kotlin, SQLite
Aug 2025 – Present · Remote
May 2025 – Jul 2025
I build Generative AI systems — RAG pipelines, LLM agents, and the serverless infra that keeps them running at scale.
AI Engineer with production experience designing and shipping Generative AI and LLM-powered applications in Python. Undergraduate from IIT Madras.
Right now I'm building production AI agents at Vsynergize AI — prompt engineering, multi-provider LLM orchestration, and RAG-style retrieval pipelines on AWS Bedrock for real enterprise workflows.
Before that, I was a Generative AI Research Intern at Annam AI (IIT Ropar), building PRAGATI-AI to help farmers improve crop yield through pollination monitoring.
Open for full-time roles, contract work and AI/ML advisory. The fastest way to reach me is email — I answer all of them.