OptiU
Meet the Opti Team

Your decision team. A roster of autonomous agents.

A team of autonomous agents — each laser-focused on one problem, each optimizing under your constraints, each pushing the optimal action into your systems. All orchestrated by Opti.

Opti — The Orchestrator
The Orchestrator

Opti runs the team.

Opti is not a chatbot. Opti is the orchestrator — the one who picks the right agent for the moment, hands off context between players, and makes sure every closed loop closes.

  • Picks the right player. For any decision the business throws at it, Opti routes to the agent built for the job.
  • Coordinates hand-offs. When one agent's output feeds another's input, Opti runs the closed loop end-to-end.
  • Talks to humans. Conversational, in plain language. Upload a spreadsheet, describe a problem — Opti parameterizes the agents and runs them.
  • Explains every move. Why this action. Why not the next-best. Auditable, override-aware.
In Opti's own words

A companion that runs optimizers, not a chatbot that writes text.

I'm Opti. I don't write essays for you. I run the optimizers, pick the best, name the tradeoffs, and push the action into your ERP, CRM, MES, or EMS. When you disagree, override me — I'll learn from that too.
Opti
How Opti shows up

Four surfaces. One orchestrator.

Same Opti everywhere — context-aware, scoped to the team in front of you.

A chat

Ask plain-English questions about the optimization in front of you. Answers come with the rationale and the alternatives considered.

A recommendation card

Embedded inside ERP, CRM, MES, EMS — surfacing the next optimal action where the work happens.

A weekly review

A calm cadence: what was decided, what was overridden, what to expect next.

An override

When you disagree, override is one click. The override event becomes training data for the next cycle.

Four kinds of player

Planners, Schedulers, Governors, Sentinels.

Every team member is a closed-loop optimizer. They sense, decide, and act — under your constraints, with the rationale attached. The four roles below say what KIND of decision each one specializes in.

Planner

Figures out the best plan. Who gets what, when, in what order — across people, inventory, capacity, money. The team's strategist.

Scheduler

Sequences live work in time. What runs next, against capacity, due dates, and dependencies. The team's traffic controller.

Governor

Holds a number on target — like a thermostat. Margin, service level, working capital. Re-balances when pressure rises.

Sentinel

Watches a signal around the clock and triggers the right response the moment it fires. The team's lookout.

And above the team: Opti — the Orchestrator. Reserved for Opti. Routes problems to the right player, coordinates hand-offs, and keeps every closed loop closing.

Featured players

Eight agents to start with.

A few of the players you'll meet across the sector teams. Each one solves a specific business problem and proves it with a number.

GovernorOptiRetail

Pricing & Margin Governor

The pricing strategist who never sleeps.

The problem

Set the right price on every SKU, every day, against elasticity, competitors, inventory pressure, and margin floors.

Result+2–8% revenue · +3–6% gross margin
PlannerOptiRetail

Smart Replenishment Orchestrator

The supply-chain quarterback.

The problem

Decide what to order, where to ship it, and which stores to transfer between — under cost, service, and shelf-life constraints.

Result+76% replenishment efficiency
SentinelOptiRetail

Shrink Sentinel

The loss-prevention watchdog.

The problem

Detect inventory shrink anomalies the moment they emerge and route the right intervention to the right store.

Result−5–12% shrink

Production Scheduling Optimizer

The shop-floor conductor.

The problem

Sequence every order across every line and shift — respecting changeovers, due dates, labor, and material readiness.

Result+5–15% throughput

Predictive Maintenance Optimizer

The asset guardian.

The problem

Decide when and where to intervene on every asset — balancing degradation signals against the real cost of downtime.

Result−10–20% downtime · −20–40% in heavy industry
PlannerOptiEnergy

Crude Optimizer

The refinery slate strategist.

The problem

Pick the crude slate that maximizes refinery margin under live unit, product, and logistics constraints.

Result+3–6% GRM uplift
SchedulerOptiCare

ED Triage Sequencer

The ED traffic controller.

The problem

Re-order the emergency department queue every minute — against acuity, capacity, and the patients who would otherwise leave without being seen.

Result−20–40% LWBS
PlannerOptiCare

Bed Assignment & ADT

The bed manager.

The problem

Place every admitted patient in the right bed against unit fit, isolation rules, and discharge pipeline — without idle beds or boarding gridlock.

Result+10–15% bed utilization · −20% boarding hours
Every player has a JD

The job description, in six fields.

You don't program these agents. You configure them. Six fields turn a generic optimizer into a player that fits your business — exactly.

01
Scope

The slice of the business this player owns.

02
Objectives

What 'optimal' actually means — the things being maximized or minimized.

03
Thresholds

When to escalate, when to alert a human, when to hold fire.

04
Constraints

The rules that must never be broken — physics, policy, contracts.

05
Dependencies

Other agents this one works with. Hand-offs are explicit.

06
Action Types

The concrete moves this agent is allowed to make in your systems.

Every player guarantees the optimal decision for the setup you give them. That's not a promise — it's what optimization means.

Spotlight JD
Job description · Governor

Pricing & Margin Governor

+2–8% revenue · +3–6% gross margin

Set the right price on every SKU, every day, against elasticity, competitors, inventory pressure, and margin floors.

Scope

All assortment, every channel, every store cluster.

Objectives
  • Maximize gross margin × revenue
  • Respect category-level margin floors
  • Avoid price wars that destroy long-term elasticity
Thresholds
  • Don't move price more than 6% in a single week
  • Flag any recommendation that drops margin under floor
Constraints
  • Competitor index bands
  • Promotional calendar
  • Inventory cover by SKU/store
Dependencies
  • Agentic Forecaster (demand signal)
  • Promo Optimizer (lift attribution)
  • Markdown Optimizer (end-of-life pricing)
Action types
  • Push new price to POS / e-commerce
  • Raise an alert when constraint nearly breaches
  • Log rationale to audit trail
Cadence

Daily — intraday on high-velocity SKUs

Where the team lives

Four sector teams. Same orchestrator.

Players are organized into sector teams, each built for the problems and constraints of that industry.

What we are not

Not copilots. Closers.

A copilot suggests. A closer decides — and pushes the action into your ERP, CRM, MES, or EMS with the rationale attached. The Opti team doesn't draft an email about your problem. It solves it, quantitatively, under your constraints, with a number you can measure.

Copilot

Drafts a recommendation in chat

Opti agent

Pushes the optimal action into the system of record

Copilot

Optimizes for words

Opti agent

Optimizes for dollars, hours, throughput, service

Copilot

Stops at the suggestion

Opti agent

Closes the loop end-to-end — and re-optimizes when conditions change

Copilot

Free-form, hard to audit

Opti agent

Explainable, override-aware, every action traced

Spin up the team — free.

Sixty seconds in the Test Drive sandbox. Run any agent against synthetic or sample data. See the action, the rationale, the lift.