Here’s a glimpse of the kinds of problems I like solving and systems I’ve worked on.
🧠 Session-Based Recommendation Engine (@bol.)
Role: Lead Engineer Built and deployed a real-time recommendation engine using memory-efficient Rust indices and Python models. Improved engagement by 15% and reduced latency to < 10ms.
📦 Sponsored Product Recommendation Indices (@bol.)
Role: Architect + Lead Engineer
Designed, built, and deployed recommendation indices for sponsored product recommendations using multiple models across both the data pipeline and serving layers, successfully replacing the legacy system and achieving target GMV before year-end.
🚀 ML Experimentation Pipeline (@bol.)
Role: Architect + Lead Engineer
Designed an end-to-end machine learning experimentation platform that allowed data scientists to go from idea to deployment in days. Powered A/B testing and metric tracking with full reproducibility.
📊 Harnessing Behavioral Data for Smarter Recommendations (@bol.)
Role: Data Scientist
Analyzed interaction patterns, session data, and content metadata to uncover user behavior insights that informed experimentation. Developed recommendation models to enhance relevance and improve overall personalization.
🧩 Personalization Platform Architecture and Engineering
Role: Architect + Lead Engineer Architected and led engineering efforts for personalization systems, including real-time streaming and bulk data pipelines. Integrated personalized banner experiences with the Pega decisioning ecosystem and models. Also responsible for secure data movement and encryption across systems supporting personalization use cases.
🌐 Custom Websites with Responsive Design and Product Showcases (@bol., @mile.bv)
Collaborated with small business clients to build responsive websites tailored to their branding and product offerings. Ensured seamless display of product information, mobile compatibility, and intuitive navigation to enhance customer engagement.
(More coming soon – this section will grow as I document past and current work.)