How AI Voice Agents Train Sales Teams at $10B Company



AI Summary

This video is an interview with Fernando Tuchi, a Senior Product Manager at Trend Micro, who discusses his work in AI and cloud security, specifically focusing on AI applications for learning, enablement, and localization. Fernando shares the story behind a voice AI platform he developed to help sales engineers and field staff practice storytelling and customer communication through simulated phone calls with AI personas. The platform provides real-time feedback, coaching, and evaluation of storytelling skills, with scenarios tailored to different personas and objectives.

Key topics covered include:

  • The origin of the AI voice persona platform as a personal language learning tool
  • Its adaptation to a business enablement tool during an internal Trend Micro AI competition
  • Features of the platform: voice-based conversations, real-time feedback, heat maps, and detailed reports
  • Challenges of designing natural and culturally appropriate AI personas
  • Importance of prompt engineering to maintain conversation relevance and naturalness
  • Use of Microsoft Azure speech services for text-to-speech synthesis
  • Considerations about latency, context window size, and costs in voice AI design
  • Current adoption of the tool in various regions and plans for broader rollout
  • Evaluation metrics focusing on skill improvement and potential business impact
  • Advice for others building AI voice agents: start with a real problem and embrace rapid technology changes

Fernando emphasizes the rapidly evolving nature of AI voice technology and the need to create engaging, human-like interactions to effectively train and enable sales teams.