Projects

Things I built to understand how they work.

Built from scratch

All 45 repos ➔

Each one started as a question about how something works underneath.

  • ★ 1miniDBin progressA B+Tree-indexed key-value storage engine built one layer at a time. Each layer fixes a problem the one below it causes: append-only log, then an in-memory hash index, fixed 4 KB pages, an LRU buffer pool with pin counts and dirty tracking, and finally the B+Tree. The append-only log layer is built and runs its own test suite (6 of 7 passing). The failing test is documented on purpose: separator-framed records desynchronise when a value contains a marker byte. The fix is length-prefixed records, the way SQLite, LevelDB and Bitcask frame them.C++17, no dependencies
  • ★ 1drinkmix-data-warehouseA star-schema data warehouse and ETL pipeline for a beverage retail chain (BTU course project). Five fact and dimension tables. Deduplicated about 47,000 records across three source systems with no shared key, with idempotent loads and count reconciliation.Oracle PL/SQL, SQL
  • PCA-Image-CompressionImage compression with PCA implemented from first principles, no PCA libraries. Splits a greyscale image into 8×8 blocks, computes the covariance matrix, finds the top eigenvectors by orthogonal iteration, and measures reconstruction quality against storage at each component count.Python, NumPy
  • Oligo binding-site retrievalresearchResearch prototype from my internship: a two-tower encoder trained on primer3 binding energies. Scored by how much of the database must be checked exactly to recover 95% of true binders. Lab code, not public. See the Research page.Python, primer3, NumPy
  • Information retrieval engineA search system built from scratch, implementing stemming algorithms and Boolean search.Python
  • ETL-API-Using-PythonExtracts records from a public API and loads them into BigQuery.Python, Google BigQuery
  • Ds-AlgoSolutions and notes from 600+ algorithm problems, more than 200 of them in C++.C++
  • Machine-LearningImplementations from ML coursework, plus Kaggle notebooks in Kaggle_Projects.Python, Jupyter
  • IPL-AnalysisInsights from the full history of the Indian Premier League.Python, pandas
  • Python_TableauAnalysis in Python and Tableau, alongside world population and blog analytics notebooks.Python, Tableau

Earlier web development

2019–2021, before I moved into data engineering.

Full-stack JavaScript apps: a MERN memories app, a React e-commerce store, a real-time chat app with Socket.io, a Twitter clone on Firebase, a blog app with its own server, a weather app using RapidAPI, a face-detection app and a Django project. All are on GitHub.