Asking Multimodal questions via your Virtual Expert
We are proceeding with the story as well:
So we built a AI Person Builder with a headless CMS on top, added Abilities to switch between different traits and behaviours.
We wanted the Agents to call different Actions, there wasnt tool calling then so we started to create something like an interpreter (later that one will be important)😅 then we found out about tool calling, or it kind of was introduces then for LLMs and what it could be used for. We implemented memory/knowledge via RAG trough the same Tactics. We implemented a Team tool so the Agents could ask each other Qiestions based on their knowledge/memories.
Story will continue in next Post!
Investment
Collaboration
Partnership