Bridging the gap between
data logic and product engineering.
I am a Product Engineer with a physics and MBA background. I don't just build dashboards—I translate complex business problems into scalable data pipelines, embed AI intelligence, and ship end-to-end applications.
The Pipeline
How I approach building products, from ambiguity to execution.
01Business Problem
Defining the core metric and operational bottleneck.
02Analytics & Data
ETL, Forecasting, and SQL/Python modeling.
03AI Integration
RAG, LLM Agents, and intelligent automation.
04Product Shipment
Full-stack Next.js or Native Swift iOS app.
Case Studies
Selected engineering projects and product architectures.
Autonomous Retail Operations Agent
An LLM-driven multi-agent system designed to optimize inventory forecasting and automate supply chain decision-making, reducing manual operational overhead.
Business Context
Supply chain managers spent 15 hours a week manually querying SQL databases to adjust inventory levels across 40+ stores based on volatile demand.
Engineering Decisions
Bypassed traditional static dashboards entirely. Built a Tool-Calling LLM Architecture using LangChain where the agent autonomously executes SQL queries, analyzes forecasting models, and posts actionable insights directly to Slack.
Tech Stack
Python, LangChain, OpenAI API, PostgreSQL, Slack API
Fitness Analytics iOS Application
A native iOS application built with Swift that processes user workout telemetry in real-time to provide dynamic, ML-powered training optimizations.
The Problem
Most fitness apps act as simple databases. Users lacked actionable insights on progressive overload and recovery optimization.
Architecture
Developed entirely in SwiftUI with a Supabase backend. Implemented a real-time sync engine utilizing PostgreSQL RLS policies to handle offline-first data caching, ensuring zero data loss during gym sessions with poor connectivity.
Impact
Achieved a 40% increase in weekly active retention by shifting from passive logging to active, predictive coaching.