Why we built from scratch

Bringing AI into the physical world.

Starting with the industries that feed it.

Today’s software is built around structured records:

ledgers in finance, patient records in healthcare, case files in law.

The physical world rarely fits that structure.

A restaurant generates critical information every minute, but most of it never reaches a database.

The next frontier of applied AI goes beyond a chat window in a system of record.

It’s systems that can perceive, reason about, and act on the physical world.

We’re building the agentic operating system for those industries, starting with restaurants.

01 / 03Perceive

Ground truth capture.

What’s on the shelf. What came off the truck. What was thrown out. The information that matters rarely arrives as a clean database record. We bring computer vision, voice, documents, web signals, and integrations together to capture what’s actually happening inside the operation.

Scattered observations resolve into a connected, three-dimensional structure: ground truth captured from the physical world.

02 / 03Reason

Prediction on stochastic data.

Demand moves with the weather, a game schedule, a nearby promotion, or a catering order. Forecasting means continuously reasoning over noisy, shifting signals across locations, so the operation can prepare for what comes next.

A field of shifting, noisy signals forms a continuous predictive surface, with a clear path through multiple possible outcomes.

03 / 03Act

Long-horizon agents that act.

Ground truth and forecasts become decisions: what to order, how much to prep, who to schedule, and which invoice to dispute. Remy carries those workflows forward every day, across every location, and adapts as the operation changes.

A branching field of possible decisions forms one connected system. A terracotta route advances through successive choices as you scroll, carrying each outcome into the next decision.

Meet the founders

Built by restaurant operators, engineers, and startup builders.

We combine deep restaurant experience with engineering and operating backgrounds from Qualcomm, Bloomberg, and Goldman Sachs.

Aakar Khanna

Aakar Khanna

Co-founder & CEO

Aakar grew up running his family’s restaurant business. He covered the restaurant sector as an investment banker at Goldman Sachs, was an early operations hire at AtoB (YC S20), and served as Head of Growth & Operations at Kouper (applied AI).

He completed his undergraduate education at the University of Michigan, where he studied Finance and Computer Science.

Arjun Chaliha

Arjun Chaliha

Co-founder & CTO

Arjun led AI and machine learning initiatives for electronic trading at Bloomberg, building systems that processed tens of millions of dollars daily. A second-time founder, he started a peer-to-peer tutoring marketplace at the University of Michigan.

He completed his undergraduate education at the University of Michigan, where he studied Computer Science within the College of Engineering.

Marble

The autonomous back of house
for restaurants.