Decagon
AI platform for building and deploying conversational agents that handle customer support across chat, email, and voice.
Updated April 2026
Overview
- Website
- decagon.ai
- Founded
- 2023
- Headquarters
- San Francisco, CA
- Segment
- Customer & Sales AI
- Posture
- AI Support Agents
Product overview
Decagon offers a unified platform to create, optimize, and scale AI agents using natural-language Agent Operating Procedures (AOPs), enabling enterprises to deliver personalized concierge experiences. Customers including Chime, Duolingo, Notion, Oura Health, Avis Budget Group, and Hunter Douglas use it to achieve up to 80% deflection rates and 65% cost reductions in support operations. It stands out with omnichannel support (chat, email, voice, SMS), user memory for context, and rapid iteration without heavy engineering.
Revenue model
Usage-based pricing via enterprise contracts: per-conversation (fee per interaction, even if escalated; ~$0.99-$1.50 est.) or per-resolution (higher fee only for fully resolved issues); median annual contract ~$386K (range $95K-$590K); volume discounts apply.
Moat
Decagon's key competitive moat is its natural language Agent Operating Procedures (AOPs), which enable non-engineering customer experience teams to build and control sophisticated, multi-step AI agents for complex workflows without professional services or heavy technical implementation, delivering superior resolution rates like 70% for clients such as Chime and Hertz. This self-service model creates high switching costs through deep integrations with internal tools and systems, proprietary operational depth for enterprise compliance via tools like Watchtower, and scale advantages from rapid deployment in tech-savvy fintech/SaaS environments, outpacing managed-service rivals like Sierra on control and precision for structured automation.