Below is the full list of parameters that can be set when creating a simulator:
1. Course or other content the simulator is based on
If you already have a course created on the platform, the simulator can be configured based on it right away.
If you do not yet have a course, you can share any relevant content (for example, documents, presentations, guidelines, or other training materials). We can convert this content into a course format on the platform, and then generate a simulator based on that course.
2. Company
By default, the simulator uses the company name specified on the platform and the company description. This helps make the scenarios more specific and relevant to your company’s industry and business context.
If you need to train in a different context, you can specify another company, or choose not to specify a company and instead describe the industry itself or a particular situational context that would be useful for the simulator.
3. Simulator language
4. AI role
You can choose one of two roles:
AI interlocutor — simulates a client, colleague, or external partner. The goal of AI is to behave like a real person in a real situation. The character does not give hints or teach, but simulates a conversation, simultaneously checking how the participant handles the dialog.
Example within the course “AI and Machine Learning in Business”: a business stakeholder approaches the participant and asks whether an AI model can be adjusted to produce more “favorable” results for management, despite data limitations or ethical concerns.
AI companion — acts as a colleague, supervisor or manager. It introduces the participant to the situation, asks questions, and may give a short comment (without hints).
Example within the course “AI and Machine Learning in Business”: a supervisor asks the participant how they would respond if asked to deploy an AI model that has not been sufficiently validated or may introduce bias into business decisions.
5. Character description
If you set a specific role, legend and tone of voice, each simulator run will reproduce the same situation. This is convenient for targeted training (for example, working with a specific objection during sales).
The level of detail can vary. For example, here is a quite detailed description but with no specific scenarios: “Helen, the head of a product team at a large bank undergoing a transformation toward a data-driven approach. She conducts a friendly discussion with the participant of hypothetical AI/ML use cases in the bank, checking how well they have understood the course theory and can relate types of ML, AI subdomains, and their limitations to real business tasks. The tone is calm and supportive, like that of a mentor-colleague, but with a focus on factual accuracy and correct terminology.”
Another example, here is a description with a specific scenario in mind: “Helen, the head of a product team at a large bank undergoing a transformation toward a data-driven approach, discusses a concrete situation with the participant. A business unit proposes using machine learning to automatically approve consumer loans based on historical client data. Elena asks the participant to walk through how they would approach this task: which type of ML would be appropriate, what data would be required, where human oversight is necessary, and what risks or limitations should be considered (bias, explainability, regulatory constraints). Throughout the discussion, Elena maintains a calm, supportive tone, asking clarifying questions and gently correcting inaccuracies, while focusing on the correct use of AI/ML terminology and alignment with real banking constraints.” In this case, there is a specific scenario and AI will reproduce similar situations.
If you set only a general role (for example, “Amanda, client”), situations will vary while maintaining this very general outline.
6. Who will take the simulations (target audience)
It is important to describe who will be participating in the simulator — for example, a frontline tech support employee, manager, account manager, sales person, HR, Product manager, etc. This information is taken from the course description on the platform but you can configure this aspect.
This affects the communication style, difficulty level, and format of the situations.
7. Simulation goal
The simulation goal is automatically created based on the course content. It defines the core measurable learning outcome and explains why the simulator is needed in the first place.
If you have specific aspects you would like to check, we can configure that.
Example: Demonstrate the ability to apply key AI and machine learning concepts to real banking business scenarios, explaining their use cases and limitations in a non-technical way.
8. Simulation objectives
Objectives break down the main goal into specific knowledge areas and skills that are assessed during the dialogue. They describe what exactly the participant must demonstrate based on the course content.
Examples:
Ability to explain how machine learning works in a non-technical, business-oriented way
Ability to distinguish between core ML approaches (classification, regression, clustering)
Awareness of AI limitations and principles of responsible AI use in banking
9. Simulation evaluation criteria
Evaluation criteria define how it is determined that a skill has been demonstrated. Evaluation criteria are automatically created based on the course content.
However, if you have specific aspects you would like to check (for example, to judge communication or adherence to instructions), we can configure them.
Within our system, Criteria can be configured in two types, depending on the simulator format. Some assess whether the participant correctly uses and explains specific course content. Others focus on how the participant communicates and acts in the situation, regardless of the specific content.
Each criterion is evaluated on a 10-point scale, and additionally an average final score is calculated.
10. Character Visual Customization
To make the simulator experience more realistic and engaging, you can customize the visual appearance of the character.
You may provide a reference photo of a real person, describe preferred appearance traits (such as age range, style, or overall look), or share an image or description of the background setting.
Based on the materials and preferences you provide, we will create a character that closely matches your expectations — either using the reference photo or generating the visual based on your description and contextual requirements.
This allows the character to better reflect your desired tone, environment, and training scenario.
Examples: