Hi cihfunlove, 

It sounds like you have a couple of questions here. 

1. There’s a question about which models power Docebo’s AI tools 
2. There’s a question about data analysis through AI (LLMs) in Docebo as it relates to the roadmap. 

Today, Docebo uses multiple LLMs under the hood to accomplish a variety of tasks. This includes models from OpenAI (ChatGPT), Anthropic and more. Check out this Support article for a breakdown of which model supports each task: https://help.docebo.com/hc/en-us/articles/22706749501842-Overview-of-artificial-intelligence-functionality-and-models-in-Docebo. For context, our team runs extensive testing on a variety of models before implementing for each task. If you see a specific model selected for a task on this table, its because our team found that model to perform best for the task. We don’t use a single model for every task, as it introduces fragility and risk choosing the wrong tool for a particular job. 

As for data analysis, I’m not aware of anything currently announced or available on the roadmap. But I’m certain that the team is already exploring this use-case, as it is quite popular and often requested. It’s a fairly complex one, insofar as data handling and LLM accuracy goes. AI is powerful, but on topics like data - it can be hit or miss. 

You can stay current on our roadmap by checking out this resource: https://community.docebo.com/ideas