AI/ML Engineer building production-grade retrieval & ML systems.
I design ML systems that bridge rigorous statistical inference and modern AI engineering — from hybrid retrieval pipelines and LLM evaluation frameworks, to predictive models with measurable business impact.
Tools I use to ship.
Projects with measurable outcomes.
AeGis-RAG — Production-grade RAG System
Hybrid retrieval (dense + BM25 with RRF fusion) with cross-encoder reranking and safe abstention to minimize hallucinations on policy QA.
Amazon ESCI Search Engine
Fine-tuned Matryoshka embeddings on the Amazon ESCI e-commerce dataset to compress vector dimensions while preserving retrieval quality.
Housing Price Prediction
End-to-end geospatial valuation pipeline using ensemble methods, with rigorous hypothesis testing across 80+ structural and geospatial variables.
Parking Vision System
Deep learning vision pipeline for autonomous parking maneuvers, with real-time spatial obstacle detection.
About.
Master’s in Applied Econometrics + Diploma in Artificial Intelligence. Eight years as an Economics & Statistics educator before transitioning into AI engineering — a path that gave me both the mathematical depth and the communication skills to turn complex ML systems into business value.
Read full bio →