About Hex: Personas are the experts, while Ensembles are the teams and

Lots of companies are using Gen AI agents to perform actions. Helpful, but you still have to know what actions to take. That’s the hard part. And doing it at scale, across thousands of players, is even harder.

That’s where Hex is different. Hex uses predictive modelling to drive the reasoning of large language models (LLMs). Through this approach, we ground every recommendation in an objective: how do we craft an offer to drive frequency, for instance, or how do we protect against this risk proactively?

Personas are the experts, while Ensembles are the teams

Through a process called fine-tuning, we have trained a network of LLMs to break down and reason through situations like different people in the casino. We call these LLMs personas, and each is an expert in her respective field.

To date, we have developed personas that think like:

  • Host

  • Casino marketing VP

  • Compliance professional

  • Surveillance operator

See more about the Hex Personas here.

Like how casinos assemble teams with diverse skills to achieve specific tasks, Hex assembles different Personas for specific use cases. We call a team of Personas organized to achieve a specific task an Ensemble. Ensemble is Hex’s application layer. The potential Ensembles are endless; explore the Ensembles we have created or are developing here.