Head of Voice AI & Speech Models

salesforce.com, inc.; Bellevue (WA); On-site; USD 197,000 - 314,000; Full time

Job summary

Salesforce is seeking a Director of Agentforce Voice Models to own the voice intelligence stack (ASR/STT, TTS, S2S) across multilingual deployments. You will partner with product, infrastructure, and GTM teams to set benchmarks for transcription accuracy, latency, and expressiveness at enterprise scale.

You will build and lead a world-class team of researchers and engineers, drive architectural decisions, and ensure responsible AI practices across voice safety, consent, and data governance.

Qualifications

• 10+ years in speech/audio ML with leadership
• Experience shipping production voice models at scale
• Strong Python proficiency and ML tooling experience

Responsibilities

• Define multi-year roadmap for Agentforce voice capabilities (ASR/STT, TTS, S2S, real-time agents)
• Set accuracy, latency, quality benchmarks and drive organization to meet them
• Evaluate build vs. buy vs. partner decisions for new voice capabilities
• Lead recruitment and development of a 20–30 person research/engineering team
• Collaborate with Product, Legal, Privacy, and Trust & Safety on ethics and compliance

Skills

Speech ML leadership

Team management

Python

Multilingual ASR

R&D strategy

Education

PhD in CS/EE or related field

Tools

PyTorch

JAX

AWS

GCP

Azure

Job description

Salesforce is seeking a Director of Agentforce Voice Models to own the voice intelligence stack (ASR/STT, TTS, S2S) across multilingual deployments. You will partner with product, infrastructure, and GTM teams to set benchmarks for transcription accuracy, latency, and expressiveness at enterprise scale.

You will build and lead a world-class team of researchers and engineers, drive architectural decisions, and ensure responsible AI practices across voice safety, consent, and data governance.