Machine learning in Video Games is not just for in-game.

Conversational AI and Narrative Generation
Generative AI for natural language has developed rapidly recently, to the point where compelling narratives can be produced with far less human effort, and personalised to the individual player and game state. We can help you leverage this technology to have non player characters respond to a much wider range of scenarios than would be possible with scripted dialogue.
Content Generation
AI can help with videogame content creation by generating, refining or optimising environments, levels, characters, along with audio and visual effects.
In-Game Non Player Character AI
While in the near future reinforcement learning (the learning used for self driving cars) will begin to play a significant role in videogames, as already demonstrated by the super impressive DOTA2 bot OpenAI Five, the hardware requirements for this type of bot to competitively play a complex real-time game are beyond those available to the gamers of today. Right now a hybrid approach works well where smaller machine learning models are combined with “traditional” game AI to make a much more realistic and unpredictable player than scripting alone, this is an area we have been actively researching.
Turn based games with their tighter set of available moves are excellent candidates for more advanced machine learning than real time games, resulting in genuinely compelling rivals in board games and other turn based games.
Natural Language Understanding & Speech Recognition
Have your players interact and influence the game through speech, just as they would with Alexa or Siri,and have natural language conversations with non player characters. Players can have all sorts of interactions with both the game world and non player character through natural language spoken conversations. For example, a non player character might provide relevant clues via natural language from a structured dataset of game data. In this way, non player characters can behave in a far more believable manner than a scripted text character.
Player Retention
The use of Machine Learning with gameplay data allows for powerful churn prediction. We will work with you to understand how best to leverage your data and determine key player metrics. We will work with you to understand how best to leverage your data and determine key player metrics. This will allow us to build predictive models which can be used both internally to prevent churn at an early stage as well as targeting specific offers to players who are likely to churn.