RXed AI News

AI news with a spine.
@RXed_EU
SCORE 7/1018/20 elements Audited 2026-07-30 · RXed table v1.0

Decagon

“The AI concierge for every customer” — the vendor’s own words

80% of Decagon's own inference now runs on models it fine-tuned in-house, and case studies show real actions, not just deflection — one OpenAI-cited customer resolves 91% of tickets with zero human involvement. But it's a $4.5B valuation against a third-party estimate of $35M ARR, roughly 130x, and reviewers keep hitting the same wall: nobody can tell you why the agent did what it did.

SpecialistAutomation & AgentsChatVoice & SpeechPaid
Vendor
Decagon AI, Inc. · decagon.ai
Origin
US — San Francisco
Pricing
Platform fee (est.) ~$50,000/yr · Usage (est.) ~$0.99/conversation or ~$0.50/resolution · Median enterprise ACV (est.) ~$386,000/yr
Users (official only)
10M+ customers served, 80% deflection rate (undated, Decagon's own site); separately, 100+ new global enterprise customers added in fiscal year 2025 (source, 2026-01-28)
Platform fee (est.)~$50,000/yrthird-party estimate; not vendor-published
Usage (est.)~$0.99/conversation or ~$0.50/resolutionnegotiated per contract; definition of 'resolution' is a documented source of billing disputes
Median enterprise ACV (est.)~$386,000/yrrange ~$95,000–$590,000+, per Vendr data via aggregator reporting

No pricing page has ever existed (Wayback Machine CDX check returns zero snapshots). Every figure here is a third-party estimate (Vendr data, aggregator reporting), not vendor-published.

checked 2026-07-30 · vendor pricing page

Element scores

Reactive
Retrieval & Memory
Orchestration
Validation
Models
Primitives
Pr9
Prompts
Em
Embeddings
Cx7.5
Context
Tr6.5
Tracing
Lg8.5
LLM
Compositions
Fc8
Function calling
Vx
Vector store
Rg8
RAG
Gr7.5
Guardrails
Mm6
Multimodal
Deployment
Ag9
Agents
Ft7
Fine-tuning
Fw8.5
Frameworks & harnesses
Ev9
Evaluations
Sm8.5
Small models
Emerging
Ma5
Multi-agent
Sy4
Synthetic data
Pc6
Protocols
In2
Interpretability
Th3.5
Thinking models
Tap or hover any element to see why it got that score.

Strengths

Agent Operating Procedures are a real natural-language-plus-code workflow layer with Git versioning, not just a marketing label, and the evaluation setup behind it — offline LLM-judge scoring plus staged online A/B tests with real statistical significance — is more rigorous than most of this category ships. The small-model strategy is paying off: over 80% of Decagon's own inference now runs on models it fine-tuned in-house, and named case studies back the resolution claims (Duolingo 80% deflection, one OpenAI-cited customer at 91% with zero human involvement).

Honest dings

There has never been a public price. The Wayback Machine shows zero snapshots of a pricing page ever existing, and third-party deal data puts the median enterprise contract at roughly $386K/year, ranging from about $95K to $590K+. Despite the Trace View and Watchtower tooling on the tin, G2 reviewers and repeated independent write-ups converge on the same complaint: operators can't always tell why the agent decided something, and audit logs and user roles are described as primitive. Agent Assist, the human-copilot feature, only works with Zendesk. The $4.5B valuation sits against a third-party ARR estimate of roughly $35M, about 130x, a number Decagon itself has not confirmed or denied.

Best for: Technical teams at high-volume support operations (15K+ tickets/month) already on Salesforce, Zendesk or Intercom, who want workflow control over chat, voice and email resolution and can resource a sales-led, multi-week implementation rather than a self-serve setup.
Visit decagon.ai/product/overview Prices and details change — this passport is re-verified at least quarterly.

Sources

Every audit lists the research it rests on — transparency and traceability are the product. Tools evolve: each audit is a snapshot of its audit date, and re-audits supersede older versions (kept below for reference).