It works like a regular chat completion endpoint: pass in a prompt and tools, get back a structured UI that users can click, fill out, or navigate.
Explainer video: https://www.youtube.com/watch?v=jHqTyXwm58c Link to docs: https://docs.thesys.dev/guides/solutions/chat
Would love feedback from anyone building LLM-powered interfaces or agentic tools.