Decision Intelligence

Optimized by math.
Validated by simulation.

The loop · how it works

From raw data to a decision that survives the real world.

Optimization shops stop at the math. Simulation shops stop at behavior. Ourus is the only one wiring them together.

  1. Step 0

    Discover the decision in your data

    Ingest the real operating environment and surface the opportunity buried in it. You do not arrive with a question. Ourus shows you the decision worth making.

    EV: grid + traffic + demographicsHospital: OR logs + staffing + acuityAirline: schedules + crew + weatherLogistics: orders + fleet + GPS
  2. Step 1

    Optimize in plain language

    The goal becomes an objective function and constraints anyone can read, solved to a provably optimal plan. Auditable and traceable, end to end.

    What's MILP?

    The math that searches billions of yes/no, where, and how-many options and finds the provably best one, not a smart guess. Airlines have used it for decades, and Ourus brings it to everyone, then stress-tests the answer against real people.

    EV: maximize coverage / $Hospital: minimize waitAirline: minimize crew costLogistics: minimize miles
  3. Step 2

    Simulate against real people

    The optimal plan meets reality. Agents behave like real people, each with a personality, a household, and a job. They respond to the plan, and the future branches when a condition changes.

    EV: households adopt?Hospital: surgeons & patientsAirline: crews accept?Logistics: drivers re-route?
  4. Step 3

    Re-solve until it holds

    The simulation's failures feed back as new constraints, and the plan re-solves. The loop closes, and the decision re-solves every time the world moves.

    EV: plan that holdsHospital: roster that holdsAirline: pairing that holdsLogistics: routes that hold
A decision, live

Watch a real decision get made. Pick the world.

The same loop runs in any domain. The common thread: decisions that must be auditable, traceable, and survive real people.

Objective

Maximize chargers reachable within a 5-min walk

Subject to
  • · ≤ feeder capacity / substation
  • · ≤ $4M budget
  • · ≥ 1 site per ward
  • · ADA access
02 · Provably optimal

Ourus returns 42 sites across 9 under-served zones. Each one is ranked and mapped, and every trade-off is inspectable.

03 · Agents pressure-test

When gas rises 20 percent, adoption jumps in four zones, and three of the optimal sites turn out to be undersized.

04 · Re-solved, holds

Ourus re-solves. It adds capacity where demand actually showed up, drops two dead sites, and stays under budget. The plan holds even when gas prices spike.

The differentiator

Everyone else simulates with probability.
Ourus simulates with people.

Traditional simulation is distributions: Poisson arrivals, Markov chains, a curve fit to the past. But decisions are not made by curves. They are made by a nurse who will not take a fourth night shift, a driver who re-orders the route, and a household weighing a car payment. Ourus gives each agent a real personality and a life, then models how those traits shape the decision. That is the difference between predicting an average and predicting a person.

Marcus, 41
2 kids · 38-min commute · rents

He will not switch to an EV until there is charging he can reach without a driveway.

Dana, 29
condo · short commute · early adopter

She switches the moment gas crosses $5, but she needs charging near work, not at home.

The Okafors
homeowners · 2 cars · suburb

They already charge at home, so a public site here is wasted spend unless a second car flips.

Branch any variable:weather / cold snapgas +20%carbon taxrecessioncompetitor openspolicy changea viral event
Why now

For the first time, frontier agents make human validation cheap, so the loop is finally possible.

Optimization has been mature for decades. What was missing was a way to test a plan against how real people behave, at low cost, before it ships. That is new, and it is what closes the loop.

Bring the loop to your hardest decision.

Contact us
contact@ourus.ai · Toronto