Headline

AI Engineer | LLM & Agentic AI Systems | RAG | ex-JPMorgan | 7+ Years in Enterprise Software | Python, AWS, MLOps | Responsible AI

About

I build production AI systems - not demos. At JPMorgan Chase, I designed and shipped a GPT-4-powered Natural Language-to-SQL platform used daily by 200+ analysts across 80+ data streams. The system combined intent classification, Retrieval-Augmented Generation (RAG), dynamic schema pruning and an evaluation framework that measured intent accuracy, SQL correctness and hallucination rate to keep every prompt and retrieval change production-safe. Before specialising in GenAI, I spent 7+ years building distributed Python systems across JPMorgan Chase, HCL and L&T. Today my focus is applied LLMs, RAG, AI agents and evaluation-driven AI systems. I strengthened this transition with an Executive MTech in Data Science while working full time and I'm now completing an MSc in Artificial Intelligence at Queen Mary University of London, specialising in Deep Learning, NLP and Computer Vision. I'm particularly interested in building AI systems that are reliable, measurable and responsible—not just impressive in demos. That means evaluation, monitoring, observability and safety are as important as model quality. I'm currently seeking AI Engineer and GenAI Engineer opportunities in London. I have full right to work in the UK and require no sponsorship. I'm always happy to connect with engineers, researchers and the London AI community.

Location

London

AI stack

antigravity

claude code

claude

gemini

github copilot

langsmith

Work experience

Software Engineer at JPMorgan Chase & Co., Aug 2023 - Sep 2025 · 2 yrs 2 mos, Mumbai, - Designed and shipped a GPT-4-powered Natural-Language-to-SQL platform enabling 200+ non-technical analysts across AMER and EMEA expense reporting to query an 80+ data-stream schema. - Reduced ad-hoc SQL turnaround from days to minutes: intent-classification agent + RAG over a curated schema/SQL catalogue + dynamic schema pruning to manage token budgets on large tables. - Built an LLM evaluation harness (golden questions, reference SQL) used as a regression gate across prompt and retrieval changes. - Designed Python financial calculators powering firmwide cost reporting across a $92B annual expense base, under strict end-of-day SLAs. - Architected an AWS ETL pipeline (S3, Glue, Lambda, SQS/SNS, Athena) modernising ingestion from 36 upstream sources.

Lead Engineer at HCL Technologies, Sep 2021 - Sep 2023 · 2 yrs 1 mo, Bengaluru, - Led design of a modular Python framework for domain-specific mathematical models, consolidating tooling and streamlining the modelling lifecycle. - Owned end-to-end data validation pipelines (XML ingestion, automated PDF reporting) across nightly batch jobs. - Defined coding standards and mentored junior engineers in an agile team.

Software Engineer at L&T Technology Services Limited, Apr 2018 - Sep 2021 · 3 yrs 6 mos, India, - Built Python module eliminating Davinci Developer ARXML tool dependency, saving ~10lakh per dongle in annual licensing costs

Education

Master's degree, Artificial Intelligence, Queen Mary University of London

Executive MTech (Master of Technology), Data Science, School of Data Science and Forecasting | DAVV University (NAAC Grade: A+)