Case Study: Five9 achieves higher self-service containment and alphanumeric transcription accuracy with Deepgram

How Five9 boosts self-service containment rates with superior speech recognition

Five9, a leading provider of cloud contact center solutions serving over 2,000 customers, needed to boost self-service success by improving speech-to-text accuracy—especially for alphanumeric inputs like order numbers and tracking IDs. To address this, Five9 integrated Deepgram’s automatic speech recognition technology (Nova-2 ASR) as an STT option in Five9 IVA Studio 7 to raise the accuracy of its Intelligent Virtual Agent platform.

Deepgram’s Nova-2 delivered 2–4× higher accuracy on alphanumeric transcriptions versus alternative STT options, enabling real-time, low-latency transcription in Five9 IVA Studio 7. As a result, a major healthcare provider doubled user authentication rates, Five9 customers saw higher self-service containment and first-contact resolution, reduced manual intervention, and measurable cost savings from shifting contacts from live agents to automated channels.

Five9

Richard Dumas

VP AI Product Strategy