THE PROOF ENGINE / 2026
I build the layer between a model's output and a decision you can defend — verified, evaluated, and cheap to run.
A LangGraph orchestration pipeline with sandboxed execution, self-repair, and 132 automated tests.
Research first. Systems second. Products that matter.
B.E. Computer Engineering, Honors in AI/ML at IIIT Pune (CGPA 8.54), topped with an executive certification in Product Management with Generative AI & Agentic AI from BITSoM — BITS School of Management.
I enjoy understanding complex problems before writing a single line of code. Whether it's reading research papers, analysing markets, or exploring how existing systems work, I believe great products begin with deep understanding rather than quick solutions.
That curiosity has led me to build production-grade AI systems spanning multi-agent architectures, LLM infrastructure, retrieval pipelines, evaluation frameworks, and developer platforms. I'm not attached to a particular technology—I enjoy finding the right approach for the problem.
My teammates know me as a problem solver who enjoys simplifying complexity and automating repetitive work. I'm currently focused on building reliable AI systems, with the long-term goal of combining engineering and product thinking to create technology that delivers meaningful business value.
Close the gap between what AI can do and what actually ships — systems that are reliable, evaluated, and worth building. The engineer half makes it work; the PM half makes it matter.
Models are easy to demo. The difficult part is making them dependable enough to earn a place in someone’s real workflow.
6 production-grade systems — each one end-to-end, with reliability, evaluation, and observability built in, not bolted on. Open any card for the architecture and the trade-offs behind it.
Product thinking, separate from the engineering: user research, problem framing, prioritisation, trade-offs, and the metrics each bet is judged on.
A confident answer is not evidence. Systems should show their work before people are asked to trust them.
Two papers published in TIJER International Research Journal — both PDFs open directly, no paywall.
Application of Long Short-Term Memory (LSTM) networks for generating musical compositions. The model learns from existing music datasets to produce new, coherent musical pieces — demonstrating the capabilities of LSTM in creative domains.
An algorithm for generating HTML code from hand-drawn web page sketches, combining YOLOv5 for element detection with OCR for text recognition to translate visual sketches into functional HTML.
From enterprise data migration at scale to agentic systems — the engineering discipline carried over.
Good AI work does not stop at the model. It reaches the edge cases, the operators, the evaluation loop, and the outcome.
The formal backbone behind the hybrid: AI/ML honors engineering plus an executive AI product management program.
Executive certification covering AI product strategy, GenAI product lifecycle, and agentic-AI product design — the PM half of the AI PM/Engineer hybrid.
Don't take the chips at face value — click any skill and the AI concierge tells you exactly where it was used.
10-agent pipelines, hybrid RAG, CI-gated evals
Click any skill — the AI answers with a concrete example, not an adjective.
Hiring for an AI role, or have a system that needs to actually ship? Available for hire — notice immediate to 2 weeks.