Cookbook: a customer support voice agent on Mercury and ElevenLabs
- Source
- x.com
- Date
🔔New cookbook: build a customer support voice agent with Mercury and @ElevenLabs. @anemenemily's walkthrough below: tool calls into a live database and streaming voice, all in one notebook. With Mercury, you don't have to choose between latency and intelligence. Diffusion LLMs give you both. Cookbook:
Context
Inception's August 28 announcement combines Mercury, ElevenLabs, database tool calls and speech. A separate current customer-support cookbook illustrates this design using Mercury 2.5 and a fictional dental clinic. It explains a concrete mechanism, without establishing what the original August notebook implemented.
ElevenLabs converts speech to text and the reply back to speech. Between those steps, Mercury selects functions for tasks such as booking an appointment. The surrounding program runs the selected function, which checks the database state, such as whether a slot is available, before changing a record. A confirmation or error returns to the model so it can formulate a reply. The availability rule lives in executable code. This example uses an in-memory SQLite database with fabricated records; it does not demonstrate deployment against real patient data.
- streaming — Sending a model's response token by token as it is generated, so the reader sees text immediately instead of waiting for the whole answer.
- AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
postInception ships Mercury Decide update with ZDR and 66k context
postInception launches Mercury Voice, a diffusion LLM for low-latency voice agents
postToday, we’re introducing Mercury 2.5, the most capable diffusion LLM on the…
postInception ships Mercury 2.5 Preview at roughly 1,100 tokens per second
Checking sign-in…
Loading comments…