Gurugram, India · Fullstack Developer × ML Engineer
Data scientist and engineer with 1+ year of production experience across enterprise financial systems and automated AI pipelines. I ship the full stack — from Java custody flows settling $1T daily at JP Morgan to LLM-powered SaaS earning ₹1.5L / month.
I don't hand ML off to another team. I build the system and the model — production-grade on both sides.
FastAPI & Spring Boot backends, React 18 / Next.js frontends, and serverless AWS data pipelines — including production Java payment flows on JP Morgan's custody platform.
LLM integration, computer vision, and applied ML shipped to production — not notebooks. Multi-modal enrichment, semantic search, and credit-risk models feeding real underwriting decisions.
I started as a systems engineer shipping enterprise backends. Scaler is how I'm deliberately closing the gap to ML engineering — trading a self-taught patchwork for rigorous foundations in data structures, system design, and applied machine learning.
DSA in Java & core system design — the CS rigour behind scalable, interview-ready engineering.
From statistical clustering and Random Forests to modern LLMs, computer vision, and vector search.
Every concept goes straight into production work — the multi-modal pipelines and risk models below.
Read it like a data flow — each stage enriched what came before.
MARXX.AI
Spinny · Kapiva · Hero · Nerolac · Banksy Museum (USA) · +15 more
Tata Consultancy Services — Client: JP Morgan
Recur Club
Side pipelines where the ideas got tested first.
Automated pipeline that parses unstructured web data in real time with Jina Reader, then runs a memory-optimized Gemma 3 architecture to extract macroeconomic sentiment and classify emerging financial risks — feeding dynamic, data-driven strategy.
An ML model assessing credit-default probability, powered by optimized SQL for time-series feature extraction over massive datasets — with comprehensive Python logging and modular architecture built to production-grade standards.
Statistical clustering over large-scale transaction datasets to segment user behaviour, plus Customer Lifetime Value forecasting to optimize marketing spend against long-term capital-growth strategy.
The product studio I build with — 38 live builds: AI SaaS products, ML systems, and brand sites across healthcare, logistics, real estate, sales, and more. Tap any card for the full case study.
What I reach for, grouped by where it lives in the system.
// let's build something
That in-between is exactly where I work best. Let's talk.