Small decisions. Interesting possibilities.Submit contentSubmit

Voice intent to tool calls

Played around with @typesafeai Jev today, mostly to understand what it does for tool calling. Instead of an LLM deciding what to do, I substituted that part with Jev. My learning is that we will be able to make the agent act before the user finishes the sentence. So far, we've used different LLM combinations…

Added to Jevfast

How it uses Jev

Input: an utterance and available tools are implied. Jev replaces the LLM decision about what to do; an agent can begin acting before the user finishes speaking, according to the experiment. Exact Jev request and answer shape are undisclosed.

Why it is interesting

The design moves tool selection into a faster decision step while leaving other LLM work in the system. The post does not enumerate tools or partial-speech safeguards.