I'm Kartik Shanbhag, a product manager at SI, formerly Sportz Interactive. I build real-time data products — and the things on this site are what I built to prove I could.
- Real-time data products
- Live event operations
- APIs at broadcast scale
- Forecasting models
shanbhag003™
Running right now
This is live, and I didn't touch it.
I built a model that plays out the rest of the Premier League season twenty thousand times and counts who won. Each bar below is a club's chance of winning the league. Arsenal at 36% means Arsenal finished top in about 36% of those simulated seasons.
The numbers are read straight from the model's own output file, which it rewrites by itself after every round of matches. If they've changed since you last looked, that's the model, not me.
At SI, formerly Sportz Interactive · 2021 to present
Five years, counted.
Client projects delivered
Feeds, APIs, dashboards and live event platforms — managed end to end, from initial requirements through to client sign-off.
Clients shipped for
Broadcasters, OTT platforms, fantasy operators, technology companies and governing bodies across India and international markets.
Auctions and drafts powered
The Indian Premier League, Women's Premier League, Pro Kabaddi League, Indian Street Premier League, The Hundred and T10 leagues — delivered live, where there's no room for a second take.
Sports covered end to end
Cricket, Football, Tennis, Kabaddi, Kho Kho, Hockey, Rugby, Curling, Volleyball, Basketball, Baseball and Golf — from data feeds and APIs to live operations, platforms and final client delivery.
Days to stand up a live auction platform
A mock auction for an Indian Premier League franchise — specified, built, tested and on screen within a week.
Sports built from scratch
Ultimate Kho Kho and Curling — built from the ground up, covering data ingestion, API design and client integration where no data feed existed before.
Clients I have worked with
Delivered through SI, 2021 to present. Broadcasters, OTT platforms, fantasy operators, tech companies and governing bodies.
Built alone, simply because I'm curious enough to build it · every repository public
Everything here is one click from the code that produced it.
Player Scout
Name a player — footballer, T20 cricketer or kabaddi raider. It finds who else plays the same way, and where they differ.
Read the case study
Premier League Supercomputer
Predicts the table, then grades itself against the bookmakers.
Read the case study
Cricket AI Digest
One live feed in. An analyst briefing and a fan briefing out.
Read the case study
FPL Auto Manager
Picks the team, submits it, and gets marked against a human.
Read the case studyHave a go yourself
Name a player. See who else plays the same way.
This is Player Scout, running live — now across football, men's and women's T20 cricket, and the Pro Kabaddi League. The football pool alone is five seasons of the big five leagues and 3,245 players, with a model that tells you where two players differ rather than just how alike they are.
Four taps to a shortlist
- 01Search a player you already have a view on, or pick one of the suggestions.
- 02Read his profile — percentile against his own position, not the whole league.
- 03Open Similar players and filter by age, contract or availability.
- 04Add anyone to the comparison and see the two biggest gaps between them.
Also here
I write about football, cricket and the systems behind them.
Five pieces, mostly written because I'd just finished building the data and wanted to know what it said. Six Tableau dashboards, two of which became articles.