## Summary so far
- Meeting started with introductions and greetings.
- Voice 3 outlined the agenda: 
  - Introduction to AI-assisted whiteboard.
  - Brief history and uniqueness of Rhyme.
  - Discussion on current voice solution challenges (using Cognity).
  - Custom demo presentation.
  - Feedback and potential next steps.
- Emphasis on open communication and questions throughout.
- Clarification sought on meeting goals: understanding reasons for meeting and challenges faced.
- Participants introduced themselves with roles:
  - Dana: Customer experience, owns metrics on satisfaction and containment.
  - Marcus: IT and systems, focuses on integration, security, and system stability.
  - Priya: Procurement, involved when there is a clear business case, cost clarity, and risk controls.
- Key client priorities highlighted: reliability, minimizing risk, especially during peak call times.
- Any change to current system must be well justified.
- Rhyme is a three-year-old company founded by Lilly Brookton Ares, with a linguistics and AI engineering background.
- Rhyme uses unique training data from thousands of hours of natural legal speech recordings rather than synthetic or podcast data.
- Rhyme powers tens of millions of calls monthly across healthcare, telecom, and enterprise support.
- Recent study of 100,000 calls showed Rhyme had the lowest hang-up rate and fastest completion rate compared to competitors.
- Client currently uses NICE with Cognity for orchestration.
- Current voice solution is stable technically but experiences latency spikes during peak times.
- Voices in current system feel robotic, leading to customer frustration, repeated requests, and increased human handoffs.
- Robotic voices negatively impact customer satisfaction (CSAT) and containment rates.
- Client has low appetite for heavy integration work unless clear return on investment is demonstrated.
- Robotic voice quality causes customers to repeat themselves or request human agents, increasing costs.
- Latency spikes during peak times increase handle times and costs.
- Client seeks clear ROI, cost predictability, and risk mitigation from any new solution.
- Success metrics include significant CSAT improvement, improved containment without increased escalations, and stable handle times.
- Rhyme’s solution is natively compatible with client orchestration, requiring minimal configuration and fast, low-cost proof of concept (POC) setup.
- Marcus emphasized no downtime or surprises allowed; deep security reviews triggered if PII is involved.
- Client wants to avoid managing multiple new vendors.
- Marcus highlighted the need for zero downtime, no surprises, and deep security reviews if PII is involved.
- Procurement requires clear cost predictability and risk mitigation.
- Operational risk is a major concern, especially for switching TTS providers.
- Tight SLAs with contractual implications for outages or latency spikes.
- On-premises, SOC 2 Type 2 hosting preferred for security and operational predictability.
- Rhyme’s model is well-run with predictable concurrency limits and GPU requirements.
- Demo of Rhyme’s AI voice assistant showed naturalistic, conversational interaction with device protection claims.
- Demo included multilingual support (English and French).
- Voice 2 noted demo voice sounded smoother and less robotic than current solution.
- Voice 2 expressed curiosity about performance at scale during peak times.
- Client stressed need for specificity and measurable success criteria, not generic AI promises.
- Next step likely to define internal success criteria before proceeding further.
- Discussion on aligning internal stakeholders for next steps including CX, platform architecture, and procurement.

## Key points
- No formal agenda slide prepared; Voice 3 plans to create one post-meeting.
- Focus on improving current voice solution experience.
- Meeting aims to gather insights on client challenges and expectations.
- Reliability and risk avoidance are top priorities for the client.
- Participants represent customer experience, IT, and procurement perspectives.
- AI whiteboard tool is being used to organize discussion and track priorities.
- Rhyme’s approach to AI voice solutions is differentiated by its linguistics foundation and unique training data.
- Rhyme’s solution demonstrates strong user engagement and satisfaction metrics.
- Client’s current voice solution uses NICE with Cognity orchestration.
- Technical stability is good but latency spikes occur during peak call times.
- Robotic voice quality leads to customer frustration, repeated inputs, increased human intervention, and lower CSAT.
- Client reluctant to undertake heavy integration without clear benefits.
- Lower CSAT drives complaints and customer churn.
- Even small increases in handle time during peak calls significantly increase costs.
- Procurement requires clear business case, cost clarity, and risk controls.
- Rhyme’s solution integrates easily with existing orchestration via API endpoint changes.
- POC setup is fast and low cost.
- IT requires zero downtime, no surprises, and thorough security review if PII is involved.
- Client prefers minimal vendor management complexity.
- Operational risk and SLA compliance are critical concerns.
- On-premises SOC 2 Type 2 hosting preferred to address security concerns.
- Predictability in concurrency and resource usage is important for operational stability.
- Demo showed natural, conversational AI voice with multilingual capability.
- Client interested in how solution performs at scale during peak periods.
- Client demands specificity and measurable success criteria beyond generic AI claims.
- Next step is to define internal success metrics before advancing.
- Internal alignment to include CX, platform architecture, and procurement stakeholders.

## Decisions
- Voice 3 to create a formal agenda slide post-meeting.
- Focus of next steps to include demonstrating clear ROI and risk mitigation.
- POC approach to emphasize minimal integration effort and quick deployment.
- Next meeting to align on scope and success criteria for POC pilot.
- Include decision makers from CX, platform architecture, and procurement in next discussions.

## Action items
- Voice 3 to prepare agenda slide after meeting.
- Rhyme team to tailor demo and POC to client’s specific business needs with minimal configuration.
- Rhyme to ensure security compliance and address PII concerns to satisfy IT requirements.
- Schedule follow-up meeting in approximately one week to discuss internal alignment and potential pilot.
- Voice 2 to send proposed times for follow-up meeting including relevant stakeholders.

## Open questions
- What specific challenges are the participants facing with their current voice solution beyond latency and voice quality?
- Are there any additional security or compliance requirements beyond PII handling that must be addressed?
- What are the preferred metrics or thresholds for acceptable latency and handle times during peak periods?
- How will Rhyme’s solution perform at scale during peak call volumes?
- What internal success criteria will the client define to evaluate the solution’s effectiveness?