Book Demo
Guided Decision Engine

A checklist lists steps. The engine routes.

The decision engine takes a situation on the floor, evaluates it against the governing standard and the manufacturer's IFU, and guides the technician to the compliant action — then captures the proof. It's the difference between a form and a decision.

The problem with a checklist is the moment the answer is "no."

A checklist records that you checked. It has nothing to say about what to do next. That gap — the space after a failed check — is exactly where compliance breaks down.

A Checklist

Lists the steps and asks you to tick them. When a check fails, it just sits there — the next move is left to memory, training, and whoever's on shift. It documents that you looked; it doesn't tell you what to do.

☐ Inspect pack for moisture
☐ Verify indicator
☐ Release load
— answer to #1 is "wet." Now what?
§The Decision Engine

Reads the answer and routes to the compliant next action, cited to the standard that governs it. A failed check isn't a dead end — it's a branch the engine already knows how to take.

Pack is wet →
ANSI/AAMI ST79: load is not sterile
Route: quarantine, do not release →
Capture record + investigate cause
The Atomic Unit

Everything is built from gates.

A gate is a single decision point: a condition, the authority that governs it, and the actions each answer routes to. Author a standard once as a set of gates, and every technician meets the same decision the same way — with the citation attached.

Where the Record Comes From
Gate Definition
ConditionIs there moisture on or in the pack after the cycle?
AuthorityANSI/AAMI ST79  ·  device IFU
If Yes →Quarantine load · investigate cause · do not release
If No →Continue to release checks
CapturesAnswer, actor, timestamp, citation — as a record

One situation in. The compliant action out.

The engine is a decision-routing state machine: it moves through gates based on the real answers, so the path always ends somewhere compliant — never in an unhandled gap.

Situation on the Floor
Wet pack found at releaseTech opens the event
Gate · ANSI/AAMI ST79
Moisture on or in the pack?Engine evaluates against the standard
▸ Moisture Present
Blocked
Quarantine load — do not release

Load is not sterile. Engine routes to containment and opens the cause investigation, both cited.

▸ Dry & Verified
Cleared
Continue to release checks

Engine advances to the next gate — indicator verification — and carries the state forward.

§
Either way, the answer, the actor, the time, and the citation become a record — the engine hands off to documentation automatically.

A checklist would have stopped at "inspect for moisture." The engine knows what "wet" means, what the standard requires, and where to send the technician next — and it proves the path it took.

The reprocessing chain is the spine.

Gates don't float free — they're anchored to where an instrument set is in its journey. The engine tracks each set along the chain, so the right gates apply at the right stage.

DecontamClean & Inspect
AssemblyPrep & Pack
§
SterilizeCycle & Verify
StorageHold & Release
Point of UseCase & Trace-Back
◉ Tracked by Set ID · one identity end to end

Each set carries a Set ID that threads the whole chain. That's what lets a decision at release trace cleanly back to the exact set, cycle, and stage it happened at.

Why it's built this way.

Deliberate, standards-driven routing — not a black box, and not a static form.

§

Deterministic, Not a Guess

The engine follows authored rules tied to standards and IFUs. The same situation routes the same way every time — auditable logic a surveyor can follow, not an opinion.

🔄

Authored Once, Current Everywhere

Update a standard or an IFU revision in one place and every gate that references it updates. No stale binders, no "which version were we on?"

It Hands Off to the Record

Routing and documentation aren't two systems. Every branch the engine takes becomes evidence as it happens. See How Records Work →

Guide → Capture → Defend

The engine guides. The record defends.

The decision engine is the first half of the platform — it makes sure the right call gets made. The second half makes sure there's permanent proof it was. Together, that's the whole loop.

See Records & Documentation

Watch a real situation route in real time.

Bring an event from your floor — a wet pack, an IUSS request, a positive BI — and see the engine take it to the compliant action, live.