Projects
Current work and past projects spanning ML, cloud architecture, and AI systems.
Startup Architecture Showcase
Production-ready reference architectures for startups on AWS — SaaS backends, real-time data pipelines, ML model serving, event-driven microservices, and full-stack web apps. Each includes IaC, CI/CD, cost analysis, and honest comparisons with alternatives.
AWS Startup Kit
A batteries-included starter template for startups on AWS. Clone it, configure it, deploy in 30 minutes. Includes API Gateway, Lambda, DynamoDB, Cognito auth, file uploads, CI/CD, and monitoring — all wired up with CDK and optimized for <$50/month.
RAG Architectures Guide
A practical guide to RAG architectures — Naive, Advanced, Modular, Agentic, and Graph RAG — with working implementations on AWS. Includes a decision matrix for choosing the right approach and cost comparisons at different scales.
Legal AI Platform – CaseRank
Built a legal AI platform that used caselaw to predict lawsuit outcomes and support consumer decision-making. Designed ETL pipelines in R to analyze legal corpus, navigated ABA regulatory compliance, and was accepted into the #1-ranked B2B Tech Wildcatters accelerator (Summer 2016 cohort).
Heuristic IPS for IoMT Devices
Lead engineer on an intrusion prevention system (IPS) for Internet of Medical Things (IoMT) devices at Johns Hopkins University. Designed heuristic detection approaches to identify anomalous behavior in resource-constrained medical hardware environments.
Real-Time Telemetry Pipeline – Microsoft
Architected a real-time telemetry pipeline on Azure Cloud to fine-tune LLM products, improving system reliability. Built prompt ranking and recommendation systems with A/B testing experiments using GenAI APIs at scale.
Dynamic Pricing ML System – AB InBev
Designed and deployed an end-to-end ML solution for dynamic pricing optimization, including data engineering pipelines, model training, and a web-based dashboard built with Flask for front-end delivery.