Say when we don't know
If the data isn't strong enough to support a move, we say so.
About Parbat
Pricing is a trust problem before it's a modeling problem.
A recommendation can look profitable on a dashboard and still fail in production. It can hurt conversion, break a feed, violate a constraint, or simply be based on weak evidence.
How the beta works
Core Belief
Your customer shouldn't be the first person to test a pricing idea.
Before a change reaches your storefront, you should know the expected upside, the uncertainty, and the constraints. After it ships, we check the projection against what actually happened — including when we were wrong.
Values
What we refuse to compromise on
The operator stays in control
Nothing ships without your approval. No silent changes, surge pricing, or per-customer pricing — and rollback is always available.
Check the work
After an approved change ships, we compare the real result with the projection and show both. That's how trust is earned over time.
Profit over proxy metrics
The decision is judged by incremental gross profit and the quality of the outcome, not activity, recommendation volume, or dashboard engagement.
Team
Two people, one product, direct line
Farrukh Zaman
Built production platforms at Cisco and Salesforce. Changing prices on a live store is a systems-trust problem before it’s a modeling problem — rollback, feed sync, failure modes at load — and that’s the work he’s done for years.
Zahara Malik
Comes from Dynata and Ipsos, where measuring how people actually respond to things — rather than what they say they’ll do — was the entire job.
We’re building Parbat because pricing is the highest-leverage decision in commerce and almost nobody has good evidence when they make it.