- Currently working on: some projects not yet ready to see the light of day. Don't fret, I got a lot of open repositories for you to dig.
- I would appreciate any feedback on my repos.
or-experiments-template — Scaffold for optimization-research group projects · Last worked on: August 2026
- Pre-commit hooks enforcing complexity limits (radon), lint/format (ruff), and stripped notebook outputs
- GitHub Actions CI (lint + pytest with coverage) gated by CODEOWNERS branch protection
- Experiment configs require a goal/hypothesis/design-note/status block for traceability
- Multiple solver backends (Pyomo/Gurobi/HiGHS, CP-SAT, heuristics) reporting into one comparable result schema
bot-detection-mlflow — Modular ML training and evaluation pipeline · Last worked on: June 2026
- MLflow experiment tracking with config-driven runs (YAML per experiment)
- Statistical comparison of algorithm performance, not just point metrics
- Model serving via a
/predictendpoint with automatic feature imputation - Health-check endpoint for deployment readiness
long-short-lstm — LSTM-based long-short trading strategy · Last worked on: May 2025
- End-to-end pipeline: data ingestion → LSTM model → trading strategy → backtesting simulation
- Dockerized for reproducible training and simulation runs
- Test suite covering ingestion, model, trading, and simulator logic
Weather-Data-Pipeline — ELT pipeline for NOAA's Integrated Surface Database · Last worked on: June 2023
- Ingests 1.4B+ hourly records (12GB, ~13,000 archive files) from 10,000+ stations via AWS S3
- Orchestrated with Apache Airflow (download → extract → combine → stage → upsert)
- PostgreSQL staging/transform layer feeding a Metabase dashboard for real-time analytics
- Fully containerized with Docker Compose
