What we need from you
CSV or Excel is fine. Your existing column names are fine. Narrative detail matters more than a tidy template. A sample review establishes whether the record has enough depth and volume to proceed.
Pricing
Start with a flashlight. Not a contract.
Four steps, each one complete on its own. Most organizations start at the free one, and decide the next move from what they see. Nobody is asked to commit to a platform before they know what their own records contain.
- Audit · free
- Diagnostic · from $4,500
- Cycle · 90 days
- Platform · scoped
Each rung stands alone. None of them requires the next.
Step one
No loginInstrument audit
Free
Ninety seconds. No data, no signup.
Fifteen questions about how you capture incidents today. You get back which contributing-condition categories your current form cannot see, and one sentence you can take into a leadership meeting.
- Coverage across all five SPECS dimensions
- The specific questions your form never asks
- A result you can screenshot or read aloud in a meeting
Step two
Most start hereOperational Incident Diagnostic
$4,500
Starting price. Fixed scope, no platform commitment.
One approved, deidentified historical export goes in. What comes back is a traceable view of the conditions that keep recurring, where the record is too thin to support a conclusion, and controls your team may want to evaluate.
- Condition map of what your record holds
- Scored patterns with named populations and denominators
- An audit of the reporting instrument itself
- Candidate controls, and the questions to evaluate them
- One live decision review with your team
Step three
The Cycle
90 days · scoped
Baseline diagnostic, then a decision review at day 90.
A baseline diagnostic opens it and a decision review closes it, which means the engagement ends by asking what the new evidence says — not with a slide deck about what you intend to do. Platform access runs for the full ninety days while the controls are live.
- Day 0
Baseline diagnostic
Your history, mapped. The starting position everything else is measured against.
- Week 2
Controls chosen
Your leadership decides which candidate controls are worth testing, and what should become visible if they work.
- Weeks 3–12
Controls live, platform running
New incidents come in as interviews. Working sessions keep the action plan honest rather than dormant.
- Day 90
Decision review
The same questions against new evidence, compared honestly — including where the answer is that nothing moved.
Step four
The platform
Scoped
By organization size and implementation needs.
Both instruments running continuously. Every new incident arrives as an interview, and every finding compounds instead of closing.
Request a platform demoWhat rates reflect
- Number of people reporting, not number of incidents — you are never penalised for reporting more.
- Implementation effort: how many source systems, how much history, how much remediation.
- Whether a diagnostic has already established your baseline, which shortens onboarding considerably.
Rates are shared in the platform conversation rather than published, because the same number would be misleading for a twelve-person service and a four-hundred-person operation.
Before money changes hands
Scope is confirmed first. Every time.
If the record cannot support useful work
We stop before the engagement begins and tell you why. A diagnostic run on a record that cannot answer the question is worse than no diagnostic — it produces a confident-looking document with nothing underneath it.
What can change the price
Larger archives, multiple source systems, and substantial data remediation can change the price or the schedule. You hear that before you commit, not in an invoice afterwards.
The current data boundary
Deidentified operational data only. chatIR does not currently accept PHI, and the contractual and technical safeguards to support it — including BAAs with applicable vendors — are still being built. Read the data boundary →
Not sure which rung you’re on?
Take the free audit. It takes ninety seconds, asks nothing of your data, and the result usually makes the next step obvious.