February 25, 2026 — Confidential

Powering the
Next Generation
of Voice

Enterprise-grade TTS built for the AI Voice Era — ultra-realistic, lightning-fast, and purpose-built to elevate every customer interaction Decagon powers.

Explore the Deck
Live in Production
Arcana v3
United Airlines IVA
"Thank you for calling United Airlines. I can help with your booking, flight status, or baggage — how can I assist you today?"
120ms
TTFB p90
1B+
Conversations
100+
Voices
10+
Languages
Today's Agenda

What We'll Cover

A focused look at how Rime elevates voice in the AI Voice Era and what a partnership looks like.

01

Introductions

Who we are and why this partnership makes sense now.
02

What is Rime?

Our models, scale, and what sets Rime apart from generic TTS.
03

Winning with Rime

Real outcomes: containment, CSAT, AHT, cost reduction.
04

The Future

PRSM, our roadmap, and the path to partnership.
Why Rime

Built Different, Built Better

Four reasons the world's most demanding enterprises choose Rime.

Scalability

Powers over 1 billion conversations globally. Infrastructure that scales instantly to your peak demand — zero warm-up, zero cold starts, zero compromises.

Pronunciation

Industry-leading accuracy for brand names, medical terminology, financial terms, and domain-specific vocabulary. No phonetic workarounds required.

Deployment Options

Cloud API, on-premise, or hybrid. Deploy in your VPC, on your own H100s, or stream from our globally distributed edge network.

Partnership

Purpose-built for Decagon. Dedicated team, co-development on custom voices, and a shared commitment to elevating every customer interaction you power.

Our Scale

Trusted at
Enterprise Scale

Built by machine learning and linguistics PhDs and backed by leading VCs — including Unusual Ventures. Rime is the voice infrastructure layer for the world's most demanding AI voice agents.

Backed by Unusual Ventures
Retail Telecom Healthcare Finance
—
Conversations Powered
—
Mist v3 TTFB p90
—
Arcana v3 TTFB p90
—
Languages Supported
Our Models

Arcana & Mist

Two models, one purpose — the most natural voice in enterprise AI.

Premium · Ultra-Realistic
Arcana v3
Studio-grade voice synthesis for brand-defining experiences. When quality is non-negotiable, Arcana delivers.
120ms
TTFB p90
H100 optimized
Maximum expressiveness — emotion, tone, cadence
Best-in-class pronunciation for any domain
10+ languages with seamless code-switching
Optimized for brand voice & intelligent IVAs
● Live in Production
Flagship Customer
Cost-Optimized · Ultra-Fast
Mist v3
Blazing-fast TTS for high-volume deployments. The fastest enterprise-grade model available — at any scale.
56ms
TTFB p90
Fastest in class
High-quality synthesis at maximum throughput
Best-in-class pronunciation for any domain
Flexible GPU — any cloud or on-prem deployment
Optimized for high-volume IVR & scale
● Available Feb 26, 2026
Flagship Customer
Results

Fortune 500 FinServ Success

Case Study — Financial Services
Switching to Rime transformed agent performance, call containment, and infrastructure costs — simultaneously.
—
Increase in sales for human agents trained with Rime's AI voice
—
Improvement in call containment — IVR to IVA with Rime voices
—
Cost reduction via on-premise Rime TTS deployment
"With Rime, our application is better than it ever was…"
— SVP Engineering, Fortune 500 Financial Services
Winning with Rime

AI Voice Era Use Cases

Superior outcomes from moving to Mist and Arcana.

Fraud Alert

Real-time, natural-sounding fraud notifications customers trust — reducing hang-ups and improving response rates.

Payment

Seamless payment confirmation and processing flows with voice that conveys confidence and authority.

Balance Info

Accurate account balance readback with correct number and currency pronunciation every time.

Loans

Complex loan term explanations delivered with the clarity and warmth that builds trust in high-stakes conversations.

↑
Higher Containment
in IVAs
+
Higher CSAT
& NPS
↓
Shorter AHT
& MTTR
↓
Fewer Escalations
to live agents
◆
Brand Trust
& compliance
Demo

Hear the Difference

Three experiences that show exactly how Rime elevates every voice AI interaction.

● Live in Production

Arcana v3 — Vespera English

The voice powering United Airlines' live voice platform. Ultra-realistic, brand-aligned, and trusted by millions of travelers every day.

0:00 —
● Mist v3 — Live

Unmatched Pronunciation

Brand names, medical terms, financial jargon — Mist v3 nails them all without phonetic workarounds. Hear Astra handle complex vocabulary naturally.

0:00 —
◈ Live Preview

ETNA

Our next-generation expressive TTS engine. Multilingual, emotionally aware, and built for the global AI Voice Era. Select a language to hear it now.

Live Waveform
0:00 —
Now playing: English ETNA · Rime Labs · 2026 Preview
Alignment for Success

Selecting the Right Use Case

Criteria for Use Case Selection

  • Multiple Areas of ImpactChoose use cases that improve multiple KPIs simultaneously — don't optimize for one metric in isolation.
  • MeasurableDefine clear before/after metrics upfront. If you can't measure it, you can't prove it.
  • DeployableEnsure the use case can go live within the 8-week window without organizational blockers.

Impact Metrics We Track

Instrumented from day one so your readout tells a clear story.

Call Containment Customer Satisfaction Task Completion Agent Escalations NPS / "Thank You" Rate Average Handle Time
Rime × Decagon · 2026
Thank You

We appreciate your time today — and we're excited about what we can build together.

Early Access · Design Partners
PRSM
Text
Speech
Tools
Not speech-to-speech

We're beginning product iteration on the next stage of Rime — and we'd love to bring Decagon along as design partners. PRSM introduces a new architectural primitive that enables speech and action to occur simultaneously. A fundamentally different architecture — and we think you'll want to be part of shaping it.