Full-Stack Web Apps
React / Vite front-ends and FastAPI / Node back-ends, wired together end-to-end with auth, APIs, and real databases.
- Responsive, accessible UIs
- REST & streaming APIs
- Auth & user management
I help founders & teams turn ideas into live, production AI products — React + FastAPI, grounded LLM/RAG, deployed on AWS. Usually in weeks, not months.
End-to-end product engineering, not just snippets.
React / Vite front-ends and FastAPI / Node back-ends, wired together end-to-end with auth, APIs, and real databases.
Grounded LLM apps that cite their sources and refuse to hallucinate — retrieval, guardrails, OCR, and streaming answers.
Ship to AWS with a real domain, HTTPS, sane security, and cost guardrails baked in — not a localhost demo.
Real, deployed products — go click around.
Problem: teams lose hours digging through PDFs, and
most AI chatbots confidently make things up.
What I built: upload your documents, ask in plain
English, and get answers with clickable source
citations — plus an honest “I don’t know” instead of a
hallucination. Hybrid search, a groundedness guardrail, scanned-PDF
OCR, per-user logins, and a spend circuit-breaker.
Result: answers you can actually act on — live on AWS.
Experiments, tools, and the source behind these products and more.
Problem: leads pile up and go cold while you
qualify them by hand.
What I built: an automation that
enriches each lead from the web,
scores it with AI, logs it to a
Google Sheet CRM (no duplicates),
emails you the hot ones with a ready-to-send reply,
and self-heals with error alerts.
Result: every lead handled in seconds, 24/7.
Submit a sample lead on the demo and watch it run.
Proof over promises — everything here is real and running.
Every project above is deployed and clickable. You can try the real thing before we ever get on a call.
Real domains, HTTPS, auth, and databases — not a localhost screenshot. I take projects all the way to live.
Grounded AI that won’t hallucinate, and infra with spend guardrails — so the result is trustworthy and the bill never surprises you.
Small loop, fast feedback, shipped at the end.
Pin down the real problem and the simplest thing that solves it.
Clear data model, clean API contract, and a UI that feels good.
Ship end-to-end in tight iterations — verified as it goes.
Live on real infra with HTTPS, auth, and cost controls.
I'm a full-stack engineer who loves turning fuzzy ideas into polished, production apps. I'm most at home where web meets AI — building grounded LLM systems that people can actually trust, with the front-end, back-end, and cloud all wired up properly.
I care about shipping: clear UX, honest behavior, and infrastructure that won't surprise you on the bill.
Pick what fits — and if you’re not sure, just ask.
A defined build — an app, an AI feature, or an automation — with a clear deliverable, timeline, and price agreed up front.
We build in tight loops when scope is still evolving — you see progress at every step and steer as we go.
A monthly retainer for new features, improvements, and keeping things running smoothly after launch.
Not sure which? Tell me about your project and I’ll recommend the simplest path.
Email me a few lines about your project. I’ll reply within a day with questions, a suggested approach, and a rough estimate — no obligation.
I’m based in India (IST) and work with clients worldwide. I keep flexible hours and overlap with most timezones for calls.
You do. On completion, all code and assets are yours — in your repositories and accounts.
Yes — that’s the whole point. I design, build, and deploy to real infrastructure with HTTPS, auth, and cost controls. Every project above is an example.
React/Vite, FastAPI/Node, Python, RAG/LLMs, vector DBs, n8n automations, and AWS (EC2, Route 53, Cognito). I pick the simplest tool that fits the job.
Tell me what you're building. I reply within a day and love a good problem.
Start a conversation →Available for freelance work · Based in India (IST), working worldwide · Replies within 24h