# Meeting summary
The meeting was a role-play sales call between Rime (represented by Colin Heilbut) and three Assurant stakeholders: Nick (Director of Conversation AI), Sarah (Engineering/Contact Center), and Mike (Contact Center). The Assurant team is evaluating Voice AI solutions for their insurance contact center, with a focus on improving voice quality, latency, pronunciation, and ease of deployment. The main outcome was a detailed discussion of pain points, success criteria, and a live demo of Rime’s capabilities, with initial alignment on key evaluation metrics and next steps for a potential pilot.

## Pain points & concerns
- **Voice quality issues**: Poor voice quality is causing low call containment, decreased CSAT/NCSAT, and poor task completion rates. (High severity; directly impacts core business metrics.)
- **Latency**: Delays in AI response negatively affect customer experience and ability to handle quick interruptions. (High severity; affects real-time interactions.)
- **Pronunciation of brand and product names**: Incorrect pronunciation undermines professionalism and customer trust. (Moderate to high severity; impacts brand perception.)
- **Handling quick interruptions**: Current solutions struggle with customers who interrupt or speak quickly, leading to poor conversational flow. (Moderate severity.)
- **Ease of deployment**: Past challenges with deploying new solutions to endpoints; need for rapid, low-friction integration. (Moderate severity; affects speed to value.)

## Success criteria for the new technology
- **CSAT/NCSAT improvement**: Measured by existing sentiment analysis tools; a significant, quantifiable increase is required.
- **Increased call containment**: Higher percentage of calls resolved without human intervention.
- **Improved task completion rates**: More successful completion of customer intents.
- **Low latency**: Fast, natural conversational response times.
- **Accurate pronunciation**: Especially for brand and product names, with customization options.
- **Seamless deployment**: Out-of-the-box compatibility with NICE and orchestration stacks; ability to deploy in hours or a day.
- **Ability to handle interruptions and multilingual support**: Demonstrated in live demo.

## Decision process & criteria
- **Stakeholders involved**: Director of Conversation AI (Nick), Engineering/Contact Center (Sarah), Contact Center (Mike); additional involvement from security and contracts teams for pilot approval.
- **Evaluation process**: Success will be measured by improvements in CSAT, containment, and task completion using existing analytics tools; pilot performance will be key.
- **Approval process**: Requires sign-off from department leads, security, and contracts. Economic buyer not explicitly identified, but Nick appears to be the champion.

## Timeline
- No explicit dates or deadlines were provided.
- Rime claims same-day deployment is possible; pilot window and urgency not specified.
- Signals of urgency include desire for rapid deployment and immediate feedback on demo performance.

## Next steps
- [ ] Colin to follow up with Assurant team to gather detailed feedback on demo performance (latency, voice quality, pronunciation). — Colin Heilbut
- [ ] Identify and involve additional stakeholders (security, contracts) for pilot planning. — Assurant team
- [ ] Define specific pilot success metrics and timeline in collaboration with Assurant. — Colin Heilbut & Assurant team
- [ ] Schedule technical deep dive on deployment and integration with NICE/orchestration stack. — Colin Heilbut
- [ ] Provide access to Rime’s voice selection tool for Assurant to evaluate voice options. — Colin Heilbut