~13 years building data and AI systems at scale
Morgan Stanley • Warner Bros. Discovery • Fox • EPAM Systems
👋 About
I build AI products and the platforms behind them — from the first user problem to the architecture review to the metrics dashboard that proves it worked.
Most AI features die between the demo and the roadmap. Not because the model was weak, but because nobody designed for retrieval quality, evaluation, cost, latency, and adoption from day one. Building for that gap — making AI a product capability instead of a prototype — is my job.
Where I spend my time:
🎯 Product direction — user problems worth solving with AI, ruthless prioritization, and honest calls on what ships vs. what stays in the lab 🏗️ Platform architecture — GenAI systems (Bedrock, Snowflake Cortex, RAG, agents), data platforms, evaluation pipelines, cost and latency budgets 🚀 Ship & scale — leading cross-functional teams from 0→1 launches through 1→n hardening: reliability, observability, and metrics that move
A lot of my production work at Morgan Stanley, WBD, Fox, and EPAM lives behind NDAs — so this profile shows my thinking, my toolkit, and what I build in the open.
🔨 What I'm Building Now
|
🤖 Enterprise AI Assistant Platform
An AWS-native conversational AI product over organizational knowledge: Bedrock AgentCore Runtime for agent orchestration Snowflake Cortex for warehouse-native intelligence RAG over Confluence + S3 with a managed vector store Deployment trade-offs benchmarked: ECS Fargate vs EKS vs Amplify Evaluation, guardrails, and cost controls designed in from day one — not bolted on |
🔭 Computational Astronomy Engine
A from-scratch astronomical computation and 3D visualization product: Swiss Ephemeris (pyswisseph) for planetary-grade positional accuracy Two rendered sky modes: symbolic chart view and true local-sky 3D projection A structured, audit-trailed interpretation pipeline — every output traceable to a finding ID Same bar I hold production AI to: deterministic inputs, explainable outputs |
🧩 How I Think About AI Products
🛠️ Toolkit
AI & GenAI
🎓 Certifications
CertificationIssuerFocusSolutions Architect — ProfessionalAWSCloud architecture at scaleGenerative AI Developer — ProfessionalAWSProduction GenAI systems on AWSSnowPro CoreSnowflakeCloud data platform & warehousingClaude Certified Architect — FoundationsAnthropicBuilding with Claude & agentic systemsPMPPMIProgram execution & delivery
📜 View certificates
🌟 Projects in the Open
|
🧮 Incometric
Income prediction from demographic and socio-economic signals — the segmentation layer behind pricing, credit, and targeting products. Python ML pipelines Model evaluation |
🎬 MovieSelect
A recommendation app built on user preference signals — recommender logic taken from notebook to usable product. Python Recommender systems Streamlit |
|
🌿 LeafCare
Plant disease diagnosis from leaf images — computer vision shipped as a decision tool, not a benchmark score. Computer vision Image classification |
🏠 HomeScope
End-to-end housing price prediction — raw data to an interactive app a buyer can actually reason with. Python Random Forest Streamlit |
📈 GitHub Analytics
🤝 Let's Talk
If you're building AI products — wrestling with retrieval quality, evaluation, agent reliability, or the demo-to-production gap — I'm always up for that conversation.
