About Me

My Journey

Building AI Engineering Foundations

My foundation in machine learning came from Brototype in Kerala — a program built around self-directed learning rather than lectures. Nobody hands you a syllabus. You find your own resources, set your own pace, and prove what you learned every single week.

That weekly review is the part that shaped me. Each one runs like a real technical interview, conducted by engineers working at FAANG-tier companies, on whatever stack or technology I had picked up that week. There is nowhere to hide in a format like that — you either understand what you built or you find out, in front of someone who does it for a living. It taught me to learn a technology fast, then defend the decisions behind it.

Those months became the discipline I still work by: collecting data, building baselines, testing models, and packaging working systems that teams and users can actually understand. It is also what took me straight into industry — first as an AI full-stack developer at Jardyan Inc, then into AI/ML engineering at StarXAI Technology.

Shipping AI Products in Industry

At Jardyan Inc, I led full-stack development of AIpublisher.org, an AI-powered journal publishing platform — building an agent-powered manuscript screening pipeline that automated quality scoring, scope matching, and formatting checks, alongside the full submission, peer review, payment, and admin workflows.

At StarXAI Technology, I designed and deployed agentic chatbot solutions for a blockchain-focused AI product: confidence scoring models for cryptocurrency news, a CAGR-based backtesting engine for trading strategies, and web-search augmentation that grounds agent answers in real-time information. Along the way I streamlined the CI/CD pipelines that carried it all to production.

Building in Public

My independent projects push the same production standards further. CorpusPilot is a multi-domain RAG platform with hybrid retrieval, mandatory citations, and a RAGAS evaluation gate wired into CI. ApplyAI orchestrates eleven agents to tailor job applications with anti-hallucination guardrails.

AuraCV, live at auracv.me, turns a resume PDF into a hosted portfolio in about twenty seconds — mine is up at robin.auracv.me. If you have a resume sitting in a folder somewhere, drop it in and you'll have an AI-built portfolio on your own subdomain in the time it takes to read this paragraph — it's free, and I'd love to see what you make. Meesho Coder is a from-scratch autonomous coding agent built on a LangGraph planner–architect–coder loop. Each one taught me something new about retrieval quality, guardrails, evaluation, and user trust.

My Current Focus

Right now I'm focused on building AI systems that are useful beyond demos: RAG products, multi-agent workflows, model deployment pipelines, and cloud-native ML solutions across AWS and GCP — and I'm open to new opportunities where that mindset matters.