I build intelligent products with LLMs and data.

Specializing in RAG, agents, prompt engineering and full-stack ML systems. I love shipping delightful, measurable experiences.

Hello!
Portrait illustration

Resume & Skills

Beyond client work, I maintain a sandbox of agents, evaluation harnesses, and creative pipelines to keep my skills sharp.

LangChain · LangGraph · Prompt Ops
Azure OpenAI · Hugging Face · Vector DBs
Python · Flask · Next.js
PyTorch · Keras · TensorFlow
Airflow · Databricks · PySpark
Azure DevOps · Docker · CI/CD
Power BI · Power Platform · Azure Analysis
  • Agent Lab
    Prototyping multi-agent flows that reason, debate, and self-heal.
    • LangGraph experiment harness with visual state playback.
    • Custom tool registry for search, policy lookup, and SQL drafting.
    • Red-team pack that stress tests hallucinations and tool misuse.
  • Evaluation Bench
    Keeping prompts and agents honest with measurable yardsticks.
    • PromptOps Lab CLI for reproducible regression suites.
    • LLM-as-a-judge plus rule-based guardrails wired into CI.
    • Latency + cost profilers for multi-model comparison charts.
  • Creative Stack
    Blending photography, audio, and narrative generation.
    • Automated photo metadata + story weaving with Lens to Language.
    • Realtime Ableton bridge for generating ambient loops per scene.
    • Microsite generator that ships zines, captions, and share cards.

Audit-Grade Experience

AB InBev · Global Risk Management
Bengaluru, India
Feb 2024 - Present
Data Scientist · IntelX Audit Copilot
  • Led a 3-5 person squad to design and deliver IntelX, an audit-grade agentic copilot running on Azure microservices.
  • Shaped GDPR-compliant retrieval across 130 markets - scrubbing PII while orchestrating LangGraph workflows over 1TB+ of evidence.
  • Onboarded 100+ auditors, sustaining 3K+ monthly questions at 99.8% evaluated accuracy and saving ~4K people-hours every year.
AB InBev · Global Procurement Audits
Bengaluru + global S2P teams
Jun 2023 - Feb 2024
Data Scientist · Source-to-Pay Analytics
  • Built GenAI anomaly detection using policy-grounded LLM reasoning on $500M+ Source-to-Pay transactions.
  • Recovered $13M in leakage spanning payment-term mismatches, early-payment costs, credit recoveries, and contract gaps.
  • Accelerated continent-spanning audits by packaging finance, policy, and operations signals into actionable briefs within days.
AB InBev · Global Inventory Analytics
Bengaluru, India
Aug 2022 - Jun 2023
Associate Data Scientist · Supply Chain
  • Automated SAP RFC → Python → Airflow → Azure Analysis Services pipelines covering $7B in global inventory.
  • Identified surplus and overbuy trends that unlocked $1B reduction and $60M working-capital recovery.
  • Delivered executive-ready KPI dashboards and forecast packs that anchored monthly SteerCo decisions.

My Projects

Side projects where I tinker with LangGraph agents, evaluation tooling, and storytelling pipelines outside of client work.

project preview

Quizzaro

GenAINext.jsAzureEducation

Instant quizzes, adaptive drills, and detailed explanations powered by Azure OpenAI to help learners improve faster.

project preview

Agentic Research Notebook

LangGraphNext.jsSQLite

A notebook app that spins up LangGraph agents to digest papers, enrich citations, and track experiment branches.

project preview

PromptOps Lab

Prompt EngineeringEvaluationsTypeScript

CLI + dashboard to stress-test prompts, guardrails, and agent flows with reproducible experiments.

project preview

Lens to Language

PhotographyCreative CodingAudio

A creative project that pairs my photos with LLM-crafted narratives, soundscapes, and shareable microsites.

project preview

Project Title

Tag OneTag Two

One‑sentence card blurb.

Why partner with me?

I help auditors and operators trust GenAI in production—shipping measurable wins while keeping governance, privacy, and uptime front and center.

4K+
Hours saved yearly
$13M
Leakage recovered
99.8%
Evaluated accuracy
130
Markets supported