Click Here to Train

PeopleSim® powering Virtual Human Role-Players

Three interconnected elements are used to simulate how real people act in difficult conversations.

Character Brains

Authentic Emotions and Personalities Power Lifelike Interactions

  • Character behavior is governed by a sophisticated virtual brain
  • Every player choice feeds into an emotional model, controlling the dynamic relationship between player and virtual character
  • Emotional state and level of trust in the player drive how the character behaves at each moment, from warm and open to hostile and distant
  • Characters have several distinct personalities that alter the emotional model from play to play, making training more nuanced and less predictable
  • Players generalize skills by learning how to succeed with different personalities

Dynamic Player Pathways

Player Choices Create Fluid Conversational Structure

  • Players navigate topics in virtually any order and choose what they want to say, when they want to say it
  • The expansive script is a map players explore freely as they find their own path to the conversation goal
  • Our unique non-branching logic allows new branches to open, close and, change depending on what’s been said
  • Seemingly infinite possible pathways allow players the freedom to change topics, ask follow-ups, and even backtrack
  • Characters remember the whole history of the conversation, just like real people

Conditional Response Lottery

Dynamic Response Weights Generate Realistic Exchanges

  • A lottery selects an appropriate character response during each conversational exchange
  • Each response is given a weighted value that sets the stage for a random selection that’s both natural and unpredictable
  • As the conversation evolves, the weighted value of each response changes dynamically
  • Players can say the exact same thing to a character in separate conversations and hear different results each time
  • High variety from play to play creates ongoing value as a training tool
Character Funnel