Summary
According to a post on X by @namanjn1998, DECIDE is a benchmark built with Bolna for the decisions voice agents make from audio. It contains 4,867 decision questions drawn from 1,081 real agent–caller clips across 32 tasks in six families: turn-taking, call routing, tool choice, language switching, caller emotion and business intent. The benchmark scores 20 direct-audio and speech-to-text-plus-LLM systems on the same questions, including cost and latency, and accepts open submissions. The post itself is mainly video, and the snippet does not fully show where its figures come from.
AI-assisted summary based on the listed source.
What happened
Built with Bolna, DECIDE has 4,867 decision questions from 1,081 real agent–caller clips across 32 tasks in six families: turn-taking, call routing, tool choice, language switching, caller emotion and business intent. It scores 20 direct-audio and speech-to-text-plus-LLM systems on the same questions, including...
Why it matters
Teams building or buying voice agents could use DECIDE to compare systems on identical questions covering routing, tool use and caller intent. Cost and latency are scored alongside decision quality, which matters for phone-based customer service. Watch whether the full methodology and the open submission process are published, since this information comes from a single social post.
Signal Intelligence
Signal Strength 91%
Technical label WATCH
Public Interest 41
Category RESEARCH
Reader Depth TECHNICAL
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
Public Interest components
Recognizable Entity Score 0
Practical Impact Score 28
Novelty Interest Score 94
Consequence Score 34
Curiosity Score 64
Shareability Score 50
Why this is here
VQV surfaced this signal because it is recent, relevant to AI Voice, connected to Scobleizer News Pulse (X).
