Worked with different path planning algorithms to best test how mobile base robots can traverse through maze-like environments!
Worked with Hetionet biomedical knowledge graph data to explore how drug compounds connect to genes and diseases, using machine learning to predict their therapeutic potential.
Outperformed OpenCV's GraphCutSeamFinder and DPSeamFinder by implementing an FMM-based seam finder for optimal image seamlines. Solved the Eikonal equation over fused cost fields (color, edge, texture, saliency) for globally optimal seam paths.
View project →
Production-ready distributed rate-limiting backend implementing token bucket and sliding window algorithms via Redis. Full audit trail persisted to PostgreSQL, HTTP 429 enforcement with per-client analytics, and deployed via Docker Swarm with GitHub Actions CI/CD.
View project →