Professional training with AI-powered dialogue simulation.
Practise in real time with realistic AI avatars that respond like genuine conversation partners. What you say is transcribed and assessed in detail, with feedback on every statement and on the conversation as a whole. If you like, add presence and voice with Video Analysis, an optional add-on module.
Seven curated scenario tracks; your own are possible at any time.
Oral exam
Practise subject questions, sharpen your structure and get used to examiners probing and to their expressions, without the pressure.
Job interview
A realistic hiring manager who engages with your actual job posting; sharpen how you present yourself.
Compliance training
Dilemma scenarios: make decisions and justify them, scalable to teams of any size.
Language training
Practise conversation (EN, ES, FR, IT, German as a foreign language); the avatar adapts to your level.
Patient consultation
History taking, informed consent, OSCE and licensing exam preparation with different patient types.
Client & leadership conversation
First legal consultations, return-to-work, feedback and conflict conversations with a realistic counterpart.
Feedback on every statement and on the whole conversation.
Immediately after the conversation: a structured overall assessment plus inline feedback on individual statements, as a dashboard and as a PDF.
The Dialogue Simulation assesses the words you say (speech is transcribed). How you come across (voice, body language, gaze) is covered by the optional add-on module Video Analysis.
When more than one person is sitting across from you.
A university viva has an examiner and an assessor, a conflict talk has a second party, a team meeting has several people at once. That is why you set the conversation format per scenario: how many characters appear, which face and name each one carries, what role it takes and when it joins in.
Tandem
Two counterparts side by side on screen the whole time. You can see how both of them react, including the one who is saying nothing. From the Enterprise tier.
Spotlight
Two counterparts in a single frame: the picture cross-fades to whoever currently has the floor. From the Enterprise tier.
Meeting
One character large, two to four more as tiles beside it; the large picture follows whoever is speaking. From the Premium tier.
The conversation format belongs to the individual scenario rather than to the scenario set, so a one-to-one talk and an exam with an assessor can sit in the same set. Tandem splits the screen and needs a window at least 900 points wide; on narrower devices the same scenario runs as a one-to-one conversation and the learner is told why. A minute of conversation counts once, even when two characters speak.
The details that make practising feel real.
Many small features that together make the difference, all of them available.
Speak instead of typing
Push to talk with live transcription as you speak, in 16 languages with automatic detection. Typing works too.
Speaking live avatar
A photorealistic counterpart (HeyGen) that looks at you, waits and follows up, rather than a text chatbot.
Microphone test & briefing
Check your microphone and read the scenario briefing before you start, so nothing goes wrong.
Self-assessment
Your own impression sits directly alongside the AI assessment and goes into the report.
Bring your own material
Document-based scenarios: upload a CV, learning material or case files, and the avatar refers to them.
Anonymous access
Credentials are generated pseudonymously; no real name is required. Data protection from the first click.
Visible data deletion
After the session you see it happen in real time: "All data deleted at OpenAI", with an audit log as evidence. Details in the privacy policy →
Access by link, QR & Moodle
A claim link or QR code, or straight from Moodle via LTI 1.3, idempotent and with no sign-up hurdle.
For teaching staff: group assessment
An anonymised AI group report and survey statistics per course; individuals appear only as "participant N".
Why it works
Built for production use.
The same platform already runs white-labelled, among others as the exam simulator of ZHL Bayreuth and as "Coach Kilian" for the German tax adviser exam. Each brand with its own logo, title and legal texts.
Frequently asked questions about the Dialogue Simulation
What is an AI conversation simulation?▾
A practice conversation with a speaking AI avatar that takes place in your browser. The avatar takes on a role such as examiner, patient or applicant, listens, responds in real time and asks follow-up questions. You try out an important conversation before it happens for real, and afterwards you receive feedback on what you said.
Which conversations can be practised?▾
Anything that works as a role play: oral exams, job interviews, patient and client consultations, sales, conflict and compliance conversations, or free language practice. In the Scenario catalogue you will find 37 public examples; new scenarios are built on request or in the studio inside the app.
Why practise with an AI avatar rather than with colleagues?▾
A human practice partner needs time, an appointment and goodwill, and by the fifth run-through their patience wears thin. The avatar, by contrast, is available around the clock: you repeat as often as you want and make mistakes without anyone watching. For institutions there is a further advantage: hundreds of learners can practise at the same time, without a per-head trainer budget.
What feedback do learners receive?▾
Two layers: feedback on individual statements directly in the transcript, and afterwards an overall assessment against criteria such as clarity, logic, precision, rhetoric and argumentation, as a dashboard and as a PDF. Optionally, Video Analysis also assesses voice, body language and gaze.
Which conversation will you practise next?
We set up your free trial access, usually within 24 hours.