Can employees run AI roleplays using our own company context and documents?

Yes. Upload your sales rubrics, product specs, HR policies and methodology, and the roleplay runs inside that context — the simulated buyer objects using your real competitive dynamics, and performance is scored against your rubric.

This is the line between a demo and a training system.

In generic roleplay, a rep practises against an invented buyer with invented objections. Useful for confidence, useless for transfer. In context-grounded roleplay, the AI buyer raises your actual pricing objection, compares you to your actual competitor, and pushes back using your real market dynamics — then scores the rep against your sales rubric.

Only the second one changes what happens on Tuesday's call.

There's also a layer above that most vendors don't reach: connecting the roleplay to the person's coaching history. If an employee set a goal of "give harder feedback without damaging the relationship" in Monday's session, Tuesday's roleplay prioritises exactly that, and Friday's summary shows whether it moved. Practice stops being an event and becomes a loop.

And when the template doesn't fit reality — which is most of the time — the learner just describes their actual situation: "I'm meeting Sarah, she's furious about the Q3 delay and wants 15% off." The AI builds the scenario from those parameters.

‍