Writing
Notes from building data pipelines, scrapers, and quantitative models — what broke, what worked, and why.
How I Scraped 20 Years of NSE Insider Filings
Backfilling two decades of insider-trade data from the NSE PIT API meant chunked date windows, composite-key dedup, and reconciling against a second source to close a reporting lag.
Building Resumable ETL Pipelines That Repair, Not Corrupt
The difference between a pipeline you babysit and one you forget about is what happens on the second run. Incremental loads, adaptive backoff, and reruns that heal gaps.
Predicting Next-Minute Price Direction on 9.4M NSE Ticks
From raw trade and quote ticks to minute bars to a directional model — and the leakage traps that make backtests lie if you let them.