Monday 07:45 · Barcelona
decision 1 of 2budget €5,000
Sarah Chen · VIP
BCN → LHR · VY8101 cancelledCritical board meeting at 14:00 in London. Cannot be postponed.
Pick an option to see how it scores.
Role-play training where the trainee actually decides under pressure, with a budget, a clock and consequences. Every choice is scored against your playbook, not against a quiz key, and the trainer gets a readiness report instead of a completion checkbox.
Critical board meeting at 14:00 in London. Cannot be postponed.
Pick an option to see how it scores.
A two-decision miniature of the real engine. The shipped simulators run 15–30 minutes with a live clock, arriving customers and a full debrief.
Same engine underneath, different playbook on top. Both are prototypes: they run, they score, they have not yet been through a year of corporate rollout.
Crisis response training for aviation and travel companies. A disruption hits, travellers queue up, and the trainee has to rebook, communicate and compensate inside a fixed budget while the clock runs.
Onboarding for delivery couriers. Learning material first, then the same material as situations: documents, vehicle check, equipment, the app, the delivery zone. An AI instructor reacts to every answer.
Neither demo is your business. The point of the call is to find the twenty minutes of your work where a wrong decision is expensive, and build that.
Talk it throughAnywhere the job is judgement under pressure and the cost of learning it live is a lost customer, a fine, or an injury.
Cancellations, outages, recalls, evacuations. The team practises the first ninety minutes, when the decisions are worst and the pressure is highest, before it happens for real.
The complaint that escalates, the refund that is not owed, the angry caller. Trainees try the wrong phrasing in a simulator instead of on your actual customer.
Courier, driver, cashier, warehouse operator, reception. New hires reach a known standard without blocking an experienced colleague for three weeks.
Data protection at the front desk, lockout-tagout, permit-to-work, anti-money-laundering checks. Every run is logged, so "we trained them" has a record behind it.
Discount pressure, procurement games, the objection nobody has a good answer to. Scored on outcome and on margin given away, not on enthusiasm.
The same scenario given to candidates and to the existing team shows where the gap actually is. Useful in hiring, useful in deciding what to train next.
Not sure which one is yours? Bring the incident that made you think about training in the first place.
Book a callE-learning asks whether the trainee remembers the rule. A simulator asks what they do when two rules collide, the budget is short and the clock is running. Those are different skills, and only one of them shows up on a bad day.
| Trainee | Largest skill gap | Avg CSAT | Severity |
|---|---|---|---|
| A. Ivanova | — | 88 | none |
| M. Petrov | Budgeting | 81 | opportunity |
| S. Dimitrov | Response speed | 64 | improve |
| K. Georgieva | Response speed | 47 | critical |
The engine exists. What takes the time is turning your actual rules into situations where the wrong choice is tempting.
What good looks like, in your words. Policies, SOPs, the incidents that went badly, and what an experienced person would have done instead.
Situations, branches, options and scoring drafted with a language model, then reviewed line by line by your subject expert. Nothing ships that they have not signed off.
A small group runs it while we watch. Options nobody picks get cut, options everyone picks get harder, wording that confused people gets fixed.
Embedded in your LMS or on its own link, with the trainer report. New scenarios added as your rules change.
Two working prototypes and no corporate rollout yet. Here is exactly what that means for you.
Both demos run end to end and score honestly, and Crisis Command already has a full debrief with skill-gap analysis behind it. What does not exist yet is the boring half: no admin panel, no single sign-on, no LMS connector, no multi-cohort reporting. Those are scoped per client rather than waiting on a shelf, and they are real work, not a checkbox.
A generic simulator teaches nothing. Yours has to encode your rules, your edge cases and your idea of a good outcome. That is why the build starts with your playbook and ends with your expert signing off every branch.
The simulator measures agreement with the standard you gave it. If your standard is wrong or unwritten, the tool will faithfully train the wrong thing. Sometimes the useful outcome of week one is discovering the playbook does not exist.
A good run is evidence that someone handled a simulated situation well. It is not a licence, an accreditation, or a legal defence on its own. Where a regulator requires certified training, this sits alongside it, not instead of it.
Use one, for the parts that are genuinely "read this and confirm". A simulator earns its cost only where the job is judgement under pressure and a wrong call is expensive. If your training problem is knowledge transfer, a course is cheaper and we will say so.
Both, deliberately. Scenarios, branches and scoring are authored and reviewed, so the simulator is deterministic and defensible. The language model is used to write the material fast and, where it helps, to play the other person in free conversation. What gets scored is never left to improvisation.
As a standalone link today, and as an embedded frame in Moodle, WebTutor or an equivalent once built — the simulators are plain web applications, so embedding is the easy part. Single sign-on and pushing results back into your LMS are scoped per client and are not built yet.
Whatever your team works in. Crisis Command is in English. Delivery Ace ships en, ru, bg and uk. Scenario text is authored per client, so the language is a choice, not a limit.
A first scenario set is a fixed-scope project, priced after the playbook intake, because the honest estimate depends on how many branches your rules actually need. After that, adding scenarios is incremental. We would rather quote one narrow module you will use than a six-module programme you will not.
Thirty minutes. Open a demo together, then find the twenty minutes of your team's work that would be worth simulating, and be honest about whether it is worth building.
Book a call with Artem