You bought the seats. Did the judgment arrive?
Adoption is not capability. Most teams have licences; few have people who delegate well, verify what comes back, and sign what ships. Miatz trains and measures exactly that — from real work, with privacy your works council can read.
The seats-versus-skill problem, in numbers
Every stat links to its source — we cite, we don't assert.
Coaching where the work happens
A coach that rides along in Claude Code, Cursor, VS Code and your chat tools — nudging at the three gates of every consequential run, and never doing the work for them.
Delegate
before the AI runs
“What exactly am I asking for — and what would tell me the answer is wrong?”
Frame the outcome, state a prediction with a confidence, bound the blast radius.
Accept
when output returns
“Did I verify this, or did it just sound right?”
Check before adopting: run it, trace the claim, reconcile against your prediction.
Deploy
before the deed
“If this ships and fails, who is affected — and did I sign it?”
Premortem consequential actions, check reversibility, disclose honestly.
Signals from these gates — verification-before-accept rates, calibration curves, gate discipline — feed each person's private Itz'at Score and your team's aggregate readiness view.
Readiness you can defend in a board deck
Team readiness rollup
Per-dimension cohort averages across delegation, description, discernment, diligence and the thinking strand — every number with a “how computed” note.
At-risk visibility
How much AI output is being shipped unverified — the cohort-level early warning for quality debt, paired with adoption context from the tools you already run.
Champions & coaching
Spot the calibrated operators worth cloning, and give everyone else a daily five-minute practice that runs on their real backlog.
Privacy by architecture
Prompts never leave the device — the schema has no field for them. Aggregates of 5+ only. An immutable consent ledger your DPO can audit.
Benchmark first
Start with a measured baseline: team calibration lab, structured interviews, and your existing Copilot/Cursor adoption metrics in one readiness report.
A credential that travels
The path ends in Itz'at AI-Fluency — observed practice plus an oral viva, issued as an Open Badges 3.0 verifiable credential.
The consent screen your people actually see
Surveillance kills the signal: people hide AI use from employers at alarming rates. Tuukul is designed so there is nothing to hide from — which is exactly why the data is honest.
Individual coaching stays private to the learner. You see the aggregate picture. That trade is what makes participation real.
Miatz Tuukul wants to coach you
This is the real consent screen — not a marketing mock.
What it reads
- ✓Signals about how you work: iteration counts, whether you verified before accepting, prediction vs outcome
- ✓Your predictions and short notes you choose to write (240-character cap, yours to delete)
- ✓Scores and streaks computed from those signals
What never leaves your machine
- ✕Your prompts or the AI's responses — the upload API has no field for them
- ✕Individual data to your employer — organizations see aggregates of 5+ people only
- ✕Anything after you hit pause — capture stops, queued events are discarded
6–8 design partners, then we close the doors
We're building the readiness benchmark and dashboards with a small founding cohort. If you want your team measured first — and a say in what the dashboards show — this is the moment.
Or write to us directly via the contact page.
What does a manager actually see?
Cohort-level aggregates only: per-dimension averages, readiness, and at-risk counts over groups of five or more. Individual learner data is visible only when that learner explicitly opts in to identified sharing. Every metric shows how it was computed.
Is this monitoring software?
No — and structurally so. Raw prompts and responses never leave the employee's machine; the ingestion schema has no field for them. What syncs is judgment signals: did they verify before accepting, how calibrated their predictions are. Coach, not cop.
How does this relate to the Itz'at credential?
The same fluency evidence rolls into each person's Itz'at Score and, over time, the Itz'at AI-Fluency credential — observed real practice plus an oral viva, not a multiple-choice test someone can braindump.
What does the pilot look like?
A founding-partner motion: we benchmark your team (calibration lab + structured interviews + your existing AI-adoption metrics), run the coaching layer with one cohort, and review the before/after readiness picture together.
See the individual experience first
The best way to evaluate the coaching layer is to feel it: run the free Calibration Lab, then explore Tuukul.