I spend most of my time in funnels — the ones users abandon and the ones products live or die by. Give me session logs, a query editor, and a drop-off number, and I'll tell you where it's happening, why, and what to ship next.
Each entry below follows the same shape: what the data showed, what I changed, and what moved. Tags mark whether the impact is measured in production or modeled from cohort assumptions.
High-value players were vanishing mid-session, and nobody knew exactly where. I pulled 20K+ gameplay sessions with SQL and Excel to find the moment.
Cleaned and validated the underlying event data first — reporting was only as trustworthy as the pipeline feeding it — which lifted data accuracy by 25% and made every downstream number defensible.
Built interactive dashboards tracking retention, session length, and revenue side by side, so product and business teams could watch the drop-off point instead of debating it.
I mapped Swiggy's onboarding funnel in Excel/SQL and found 35%+ of users dropping off during the login and location-permission steps — right before first order.
Redesigned the flow around guest browsing, cutting the path from 5 steps to 3, using the funnel data and UX heuristics to decide exactly which steps to remove rather than trim arbitrarily.
10K+ transactions held a story about which categories converted and when — but no one had segmented it by time.
Built time-based demand models in SQL/Excel to flag peak conversion windows, then used them to steer inventory allocation and dynamic pricing toward the moments that mattered.
A full spec written for the Swiggy onboarding redesign — the format I'd use to hand a case study off to engineering. Fields inferred beyond what the underlying data supports are flagged rather than presented as fact.
Reduce onboarding friction and increase the share of new users who complete onboarding and place a first order.
New users installing the app for the first time, particularly those dropping off at login or location-permission.
My work has consistently centered on using data to understand user behavior and improve outcomes. During my internship, I analyzed gameplay data and built dashboards that shaped product decisions, and my independent case studies came from the same instinct: find where users get stuck, then figure out what to change. Over time I noticed I wasn't satisfied just naming the problem — I wanted to decide what got built to fix it. Product management is that combination: analytical thinking, user empathy, and cross-functional decision-making, aimed at outcomes that matter for both users and the business.