The story behind Koovis AI.
Koovis AI is a film studio. Under the name Koovis Studios we make our own films, shorts and series, written from scratch. AI lets a small team put stories on screen that would otherwise need a full crew and a large budget.
We don't make films for clients. Every story is ours, and so is the responsibility for how it's told.
People write the scripts and perform the scenes. Several AI models render the shots around those performances. A sound designer and a composer finish each film.
What we're building toward
Films that people remember for the story, not for how they were made. Starting with short films and a micro-series, and growing into longer work as the craft and the tools allow.
Raj Kolachana
Founder · Writer, director, actor
The path here wasn't straight
I started in structural engineering — IIT Roorkee for my B.Tech, then IISc Bangalore for my M.Tech. Scored GATE AIR 5 (top 0.013% nationally), which in Indian engineering circles opens every door. But somewhere between finite element analysis and optimization theory, I realized the same mathematical frameworks that model bridges and buildings could model human behavior, markets, and decisions. That pivot changed everything.
My first real data science role was at InMobi, where I won the Rising Star Award and scaled an ad account from $3K to $80K in daily spend. Then AgreeYa Solutions, building pricing optimization models for Best Buy, Sam's Club, and Dick's Sporting Goods. Both taught me what production ML actually looks like — messy data, tight deadlines, and systems that have to work at 2 AM on a Saturday.
Seven years at Amazon
Amazon is where I learned what it means to build at scale. Seven years as a Senior Data Scientist, shipping ML systems across global marketplaces — recommendation engines, NLP-driven review ranking, Bayesian reorder models, paid- advertising optimization, and an NL-to-SQL tool that went from hackathon project to a production tool used by thousands of account managers worldwide.
The real takeaway wasn't the scale — it was the discipline. The operational rigor. The understanding that a model is maybe 20% of a production ML system; the rest is pipelines, monitoring, failover, and the boring engineering that keeps things running at 3 AM.
Why I left
After 11 years in the industry, building systems for other people's products, I wanted to make my own work. Film is where I landed: AI finally makes it possible to tell a story on screen without a studio behind you, and the engineering discipline turns out to matter as much on a film as in production ML.
I write, direct and act in Koovis films. Actors, a sound designer and a composer join each production. Koovis stays small on purpose: the story decides what each film needs.
It's early. Our first short film is in development. Ask me again when it's out.
What we believe.
Story First
Every film starts with an original script. If the story doesn’t work on the page, no amount of rendering will save it.
Real Performances
People act every scene. The performance is filmed for real and leads the shots we generate around it.
Honest About AI
Every release says how it was made. Everyone who appears on screen has signed a release.
Restraint
Hard subjects are handled with care. We show consequences, not spectacle, and take advice from people who know the subject.