Selected work

Portfolio & Technical Milestones

Below is a highly curated list of machine learning achievements and technical artifacts that demonstrate my ability to model and scale complex data architectures.

  • 2025

    No. 01

    Kaggle · Competitive ML

    Top 60 Globally out of 150,000+ Competitors

    Achieved a peak global ranking of 60th among all registered data scientists worldwide on the Kaggle platform. Consistently delivered high-accuracy models across diverse tabular, image, and text datasets through expert ensemble techniques and advanced feature engineering.

    Duration Ongoing
  • 2022–2025

    No. 02

    Top-Tier Vertical E-Commerce · Core Services

    Core ML Systems at Scale

    Led a small, hyper-focused team to design, build, and scale core machine learning components. Optimized real-time ranking algorithms boosting search conversion by 14%+, architected high-throughput microservices handling millions of daily inference requests, and designed a custom feature store reducing training/inference skew to zero.

    Duration Multi-year
  • 2024

    No. 03

    Personal Project · Open Source

    Ultra-Lightweight Feature Store

    Developed an in-memory, zero-dependency feature storage framework optimized for high-throughput, low-latency batch offline testing and online inference parity. Built with Rust, Python, and Redis.

    Duration 12 weeks
  • 2024

    No. 04

    Personal Project · Open Source

    Robust Covariate Shift Detector

    An automated statistical tool designed to identify and alert engineering teams of drift in complex feature distributions prior to system performance degradation. Built with PyTorch, SciPy, and Pandas.

    Duration 8 weeks

References

References available on request.

Most of these engagements included a reference call at the end. If a particular case resonates, I'll ask the client before sharing their name — and I'll always let you know if I can or can't.