Voice Agents

Routines come back whole, on the device.

SEE A RUN>
memorable
$ memorable recall "run the movie night routine"
0.880 procedures/2f6b8d17-movie-night [lexical,semantic]
$ memorable show procedures/2f6b8d17-movie-night
Decisive steps last time:
1. [execute] Dim: living room lights

A Voice Agent Has No Spinner To Hide Behind.

Dead air

Reasoning a routine out step by step is the wait the person hears.

Taught last week

A routine comes back whole, in about 60 milliseconds.

Never spoken

Sensitivity gates at delivery. The agent can know things it never says.

Hear, Replay, Log.

Memorable keeps the path that worked and hands it back before the next run.

The request is matched to a stored routine before the model starts reasoning.

The routine runs whole: lights, volume, the film. Nothing is re-explained.

The outcome is stored. A routine that changed is kept as a revision beside the old one.

memorable
$ memorable recall "run the movie night routine"
0.88 procedures/2f6b8d17-movie-night

How It Works For Voice Agents.

Ability traces come in at the end of each run. On OpenHome, 100% task completion with procedural memory, up from 70%.

{ 01 }

Routines come back whole

A routine taught in a past session replays as one procedure, not a re-derived plan.

"movie night"spokenroutineone procedurelightsvolumefilm

What Memorable Reads.

2 CONFIGURED
Ability traces from the harness
One trace per triggered ability, joined to its result
Any voice harness
POST the steps to /v1/extract and get the procedure back

Who This Is For.

Voice platform engineers

Own the harness and want the gap between asking and hearing to shrink.

Ability builders

Ship multi-step routines without writing every step into the prompt.

Product leads

Watch completion and latency on repeat requests.

More Use Cases.

( LET'S TALK )

Let's Build Something
Memorable

Book a Call