Work
A focused selection of backend systems, applied ML, and data engineering. Each case study explains the contribution, constraints, and outcome.
Software Engineering
Active development
Full-stack support operations combining dependable backend workflows, lifecycle-aware analytics, and advisory routing suggestions.
ContributionBackend workflows · Integration safety · Operational analytics · Model-release tooling
10,803 / 96.7% Current ticket corpus / records carrying support-group IDs
FastAPI · PostgreSQL · React · TypeScript
Research
Princeton appointment concluded · May 2026
Multi-vendor speech and translation evaluation followed by a recoverable end-to-end pipeline for parliamentary records, campaign audio, and structured analysis data.
ContributionTool benchmarking at Studio Lab · End-to-end data-pipeline engineering at Princeton
40+ GB 2014 and 2019 campaign audio processed after multi-vendor tool evaluation
Python · Selenium · SQLite · AWS
Software Engineering
Private team project
Team-built full-stack Flask application using locally run computer vision and primarily open-source application, OCR, and data frameworks to retrieve race photos by bib or selfie.
ContributionMatch Verification · GPX Race Creation · Upload/Search UX · OCR/SWT Evaluation
32 commits · 7 merged PRs My mainline contribution across 24 paths, including verification, race creation, and athlete-facing workflows
Flask · PostgreSQL + pgvector · InsightFace / ArcFace · PaddleOCR