Product ownership · Hands-on delivery
Finding user friction.
Building better products.
Product Manager with 6+ years of experience across product management and consulting, including 3+ years owning and delivering products.
I own product direction for Media.net’s analytics platform and independently build consumer products around real user needs.
- 500+
- Daily active platform users
- 40%
- Fewer support requests MoM after Phase 1
- 1–2s
- Query times for targeted cached patterns, down from 6–10s
What I do
- Own Media.net’s platform direction — 500+ daily active users across account managers, business teams & publishers
- Phase 1 UI redesign followed by a 40% month-over-month drop in support requests
- Query failure rate cut from 1.2% to 0.3%
- IPL Strategist — probabilistic match prediction engine, live
- HumSaath — primary expense tool for 6 travellers, average age 62
- Workout Tracker — used intermittently by 14 group members
- Two-phase UI roadmap: frequently used workflows first, admin capabilities next
- Targeted caching cut query times from 6–10s to 1–2s
- Productised internal workflows adopted firm-wide at ZS
- Requirements synthesis across business units & publisher clients
- Led an 8-person engineering & QA team through the UI redesign
- Cross-functional ops, sales & logistics coordination at A&M
Experience & products · select to explore
Workout Tracker
IPL Strategist
Multi-Sports
Movie Industry Explorer
Associate Product Manager · May 2023 – Dec 2025
Bengaluru, IN
- Owned the visualisation drawer overhaul and UI revamp; led an 8-person engineering and QA team through a two-phase roadmap, prioritising frequently used features while maintaining legacy access. Phase 1 is live; Phase 2 is nearing launch.
- Introduced chart presets so users can start analysis without manual setup. Following Phase 1, support requests fell 40% month-over-month, with users reporting easier task completion.
- Identified frequently queried patterns in pipelines processing ~90GB daily and selected dimensions and metrics for targeted cache tables; reduced query times from 6–10 seconds to 1–2 seconds for those patterns.
- Conceived Compare Period after observing users export reports for manual comparisons. The in-platform feature is now the first visualisation in Black Friday dashboards.
- Delivered 20+ dashboards consolidating business lifecycle metrics into one view, replacing analysis spread across reports; each sees ~3–4 daily active users from business teams.
- Partnered with engineering to reduce query failures from 1.2% to 0.3%, a 75% improvement.
- Supported post-merger integration for 5 products acquired by Dr. Reddy's Labs; helped achieve 102% of a $2.6M monthly target across India.
- Built 7 real-time dashboards monitoring sales, coverage, and inventory across 2000+ towns, enabling tactical decision-making.
- Developed monitoring tools to track daily stock levels and in-transit inventory across 2500+ distributors, reducing stockout risk.
- Coordinated inventory planning for 20+ SKUs across 10+ warehouses; worked cross-functionally with ops, sales, and logistics teams to streamline replenishment processes.
- Supported sales operations and GTM analytics for $100M+ pharma brands; backed $3B portfolio decisions with market landscape insights.
- Built internal analytics products and reusable tooling to drive firm-wide productivity, including sales force sizing, alignment, and placement strategies for global pharma clients.
- Productised internal workflows — migrated legacy processes to Python/R, reducing processing time by 83%.
- Managed 100GB+ datasets, built reusable analytics modules, and collaborated cross-functionally with engineering and product teams.
- Owned pricing, demos, and commercialisation support for internal analytics assets adopted across global client teams.
Splitwise-style apps are too fiddly for parents managing shared spend across countries — tiny text, mental math, and no clear record of who owes whom.
Built a shared-expense product used by 6 travellers (average age 62) as their primary expense log and splitting tool during a 9-day trip. Large tap targets and INR/IDR support make shared spending easier to manage.
- Iterated to add an admin view for editing existing transactions.
- Automated who-owes-whom calculations and cash settlements, backed by Google Sheets.
Posting workout photos and manually copying and editing a WhatsApp leaderboard led to missed updates. One organiser had to reconcile photos and scores from the chat.
Built a day-level workout tracker used intermittently by 14 group members, with one-click navigation to WhatsApp to share photo and leaderboard updates while keeping the group’s existing sharing habit.
- Daily logging, photo proof, leaderboards, and group and individual progress views.
- WhatsApp leaderboard sync supports updates made in the existing group chat.
Live stats for cricket, football, F1, and tennis live in separate single-sport apps — there's no one screen to track matches across sports at once.
Compose a custom live dashboard across 4 sports — cricket, football, F1, and tennis — with a 3–6 widget resizable layout and zero API keys required. Bind each widget to one match and one stat; widgets auto-refresh while play is live.
- Every data source maps to one normalized match shape, so a single set of stat renderers works across all four sports.
- Shared server-side caching — all viewers share one upstream hit per interval per match; no keys or secrets required.
- Live-gated polling — widgets refresh fast only while their match is live, and slow-poll otherwise to catch kickoff.
Pre-match cricket predictions are usually gut calls — commentary and punditry with no visibility into the historical matchups actually driving the numbers.
Probabilistic predictor trained on 9 seasons of ball-by-ball data, built on a custom Encounter Index algorithm across 5 adjustable prediction pillars — with matchup heatmaps and a transparency layer explaining every prediction. Correctly predicted 45 of 68 matches (66%) in walk-forward backtesting for 28 Mar–24 May 2026.
- Designed the Encounter Index (EI) — weights historical player matchups by over-by-over occupancy distributions, adjusted dynamically for toss state (batting first vs. chasing).
- Interactive dashboard with 5 pillars: H2H, Matchups, Venue, Form, Player vs. Team.
- Validated via walk-forward regression and backtesting against 2026 season fixtures.
Comparing box office numbers across decades is misleading without adjusting for inflation — a $50M budget in 1995 isn't a $50M budget today.
Interactive dashboard surfacing global film industry trends across countries and decades — joining TMDB film metadata with CPI inflation data across 3 lenses: budget, revenue, and ratings.
- Joined TMDB movie metadata with CPI series to normalise budgets and revenues across eras.
- Slice by country and decade to compare industries on budget, revenue, and IMDb ratings.