About Data-Hat AI
We built the system we wished existed.
Data-Hat AI is led by Kshitij Kumar (KK), former Chief Data Officer at Farfetch and Zalando. We spent years on the other side of the table, running demand, allocation and data for global retailers, and kept hitting the same wall.
Origin
Why we started
Between us we've run planning, allocation and data for retailers moving billions of euros of stock through thousands of stores. The tooling was never the constraint, the enterprise stack recorded everything that had already happened, in perfect detail, far too late.
What no system could tell us was simpler and far more valuable: what the store wanted and never got. That answer lived with associates, in conversations, on the shop floor - and it evaporated at closing time.
Data-Hat AI exists to capture that signal and turn it into a decision. Not another dashboard. A layer that sits above what you already run and tells your teams what to move, where, and by when.
"Every retailer we've worked with was optimising the wrong half of the problem, the stock they sold, not the demand they missed."
Kshitij Kumar · Founder
Leadership
Built by people who ran this at scale.
Orkestra is designed by operators, not by a team guessing at retail from the outside. Every default in the product comes from a decision someone here had to make with real inventory on the line.
Kshitij Kumar
Founder · Data-Hat AI
Chief Data Officer
Chief Data Officer
What we believe
Four positions we build from.
- 01
Retail doesn't have a data problem. It has a decision problem.
Most retailers already sit on more data than they can act on. The gap is between the dashboard and the decision, who moves which unit, to which store, today.
- 02
Demand you never served is still demand.
Sales data only records what you happened to have on the shelf. The size that walked out, the shade never restocked, that signal is invisible to the stack and it compounds every week.
- 03
Every store is its own market.
A national forecast averages away the thing that matters. Orkestra builds a demand profile per store, not per region, because the shelf is where the money is won or lost.
- 04
A decision layer sits above the stack, never replaces it.
No re-platforming, no migration project. We read your existing POS, ERP and WMS feeds and hand back decisions your teams can execute the same day.
Want to see Orkestra on your own stores?
A 48-hour pilot on your existing feeds, no re-platforming.