What I learned from analyzing live-commerce growth.
Category mix, show timing, and inventory concentration — what per-show profitability actually revealed.
read · coming soonData engineer with 3 years on Snowflake pipelines, SQL validation, and production data workflows. Now doing Northeastern's MS in Business Analytics — building toward product and analytics roles in commerce, sports, marketplaces, and consumer products.
I grew up between two obsessions — cricket and computers. The first paid off early: Team USA U19 in 2019, and the youngest domestic player taken in Minor League Cricket's inaugural draft. The second is the longer arc, and the one I'm doubling down on now.
At Rutgers I studied Computer Science with a minor in Entrepreneurship. The CS gave me the tools; the entrepreneurship minor gave me a reason to use them on real problems. The first one was a friend's sneaker drop community — about 350 members — which became my first analytics case study: a Google Sheets tracker for engagement, retention, and inventory patterns.
After graduation I joined Bahwan Cybertek, a Snowflake premier services partner. Two and a half years on production migrations, validation, and KPI dashboards, then six months at Flex Consulting Group building a four-pipeline chain for a healthcare provider. Solid engineering. But I kept circling back to the questions the data could answer, not just the pipes that moved it.
That's the pivot. I'm now in Northeastern's MS in Business Analytics, graduating December 2027. Alongside it I build GameTime Picks, a free five-sport analytics platform that grades every forecast it publishes against official results, and I keep the P&L model behind a friend-operated live-commerce business that put per-show profitability in front of its sellers for the first time.
The plan: a Summer 2027 internship at a sports, betting, marketplace, or consumer-product company that takes data seriously and ships fast. Long-term: be the person on a product team who can both read the data and understand the customer.
Capability framing rather than role framing — these are the things I can already do well, and the lens I bring to a team.
Production ELT into Snowflake with reconciliation and lineage built in from day one. 10+ flows live at Bahwan, validation cuts post-load issues by ~20%.
Looking at messy data and finding the question worth answering — category mix, retention cohorts, profit-per-hour, growth driver patterns. Python + pandas in notebooks.
KPI views leadership actually checks. Built weekly performance dashboards from platform logs at Bahwan that cut data issue detection cycles by ~35%.
Reading data and customer behavior at the same time. Drawn to consumer, sports, live commerce, and marketplace problems where engagement signals are as interesting as financials.
Three project contexts where I apply data, product, and analytical thinking. A five-sport analytics platform I built and run, which grades every forecast it publishes. A friend-operated live-commerce business whose per-show profitability I model. And a friend-run sneaker community studied through a lightweight tracker.
The P&L and stream-level performance model for a friend-operated sports card business on Whatnot. It gave three sellers their first view of per-show profitability across $160K GMV, 12,470 orders and 479 live hours, and found 23 of 69 streams running at a loss.
A product and analytics case study focused on sneaker release behavior, community engagement, inventory tracking, and retention signals using a lightweight tracker and analysis framework.
A free five-sport analytics platform I built and run, covering MLB, NFL, NBA, UFC and EPL. Python pipelines and GitHub Actions publish, score and settle every game without manual intervention, and a settlement layer grades every projection against official box scores, so the misses stay visible next to the hits. This year I released Ask GameTime, an LLM assistant answering questions from the platform's own 14 tools, with a deterministic verifier that rejects any number absent from the source data. Simulation-first and educational. No payments, no subscriptions, no ads.
Hover any tool to see which methods it powers. The connections matter more than the inventory.
What I'm actually working through right now, alongside the MSBA coursework.
Short writeups I'm drafting on the way to publish. Three at a time, no more — the queue stays honest.
Category mix, show timing, and inventory concentration — what per-show profitability actually revealed.
read · coming soonMost "the dashboard is wrong" tickets are actually upstream data tickets.
read · coming soonWhen the more interesting work is downstream of where you're sitting.
read · coming soonCricket isn't a hobby footnote — it's how I've spent half my life. The discipline transfers more than I thought it would: preparation, calm under pressure, reading patterns in real time. All things product analytics rewards.