Rankings updated monthly — View live standings →

Decagon

AI customer service platform valued at $4.5B, pairing strong action-taking with a growing Voice 2.0 phone agent.

8.3
AT500 Score
Rank #10 in Customer Support AI
▲ +4
Starting price~$433K/year median (custom)
Pricing modelEnterprise
G2 rating4.8/5 (40)
Founded2022
HeadquartersSan Francisco, CA

About Decagon

Decagon is an AI customer service platform aimed at high-volume consumer and B2B brands, focused on conversational automation, analytics, and natural-language workflow control that appeals particularly to tech companies with engineering resources to back configuration. It tripled its valuation to $4.5 billion in under six months by January 2026, and Vendr-observed contract data puts median deal size around $433,000. In spring 2026 Decagon launched Voice 2.0, adding interruption handling, custom tone controls, branded caller IDs, and proactive outbound voice, integrating through Amazon Connect, RingCentral, and SIP trunking, though G2 reviewers note voice capability is still evolving relative to its more mature chat and email backbone.

Key features

Voice 2.0
Interruption handling, custom tone, branded caller ID, and outbound voice, launched spring 2026
Natural-language workflows
Engineers can configure agent logic in plain language
Knowledge Suggestions
Proactively identifies gaps in the underlying knowledge base
Watchtower
Monitors AI response quality in real time with built-in A/B testing

Pros and cons

✓ Pros
Product philosophy goes beyond deflection, actively improving the knowledge system over time
Strong natural-language workflow control appeals to engineering-savvy teams
Real-time quality monitoring via Watchtower and built-in A/B testing
Chat and email automation is mature and well-proven in production
✗ Cons
Cost is prohibitive for most teams, with median contracts around $433,000
Voice capabilities are newer and described by reviewers as still evolving
No self-serve option; fully custom enterprise sales engagement required
Best suited to tech companies with engineers to back configuration, less turnkey than self-serve competitors

Best for

Tech companies doing primarily chat and email supportEnterprises wanting natural-language workflow control with engineering supportTeams focused on continuously improving support knowledge qualityCompanies evaluating alongside Sierra for enterprise AI support

Why this score

AI support automation startup with strong Series A traction in 2026

G2 4.8/5 early. Series A traction 2026.