Decagon

Autonomous AI agents for enterprise customer support

Decagon builds AI agents that handle customer support conversations across chat, email, voice, and SMS for large enterprises. Rather than relying on scripted chatbot flows, the platform uses what it calls Agent Operating Procedures (AOPs) — natural-language workflow definitions that let non-technical support teams shape agent behavior while engineers retain control over integrations and guardrails. It connects to existing systems like Zendesk, Salesforce, and Stripe so agents can complete real actions (refunds, account updates, cancellations) rather than just answer questions. It’s built for high support-volume enterprises, with customers spanning consumer and SaaS brands. Its main differentiator versus simpler chatbot vendors is this combination of action-taking, omnichannel coverage, and a workflow-authoring layer designed for support operations teams rather than developers alone.

Compliance

GDPR ISO 27001 SOC 2

Visit Decagon

Key Features

Agent Operating Procedures (AOPs): Natural-language workflow definitions that let support staff design and adjust agent behavior without code, while engineers retain control of integrations and guardrails.
Omnichannel deployment: Agents operate across chat, email, voice, and SMS with shared context, so conversations can move between channels without the customer repeating themselves.
Action execution: Integrations with tools like Zendesk, Salesforce, and Stripe let agents complete tasks (refunds, account changes, cancellations) instead of only generating text replies.
Multilingual support: Policies written in one language can serve customers in many languages without maintaining separate per-language knowledge bases.
Watchtower monitoring: A built-in oversight layer that flags low-confidence or risky agent responses for review, intended to catch errors before they reach customers.
Pre-built enterprise integrations: Connectors to common CX and business systems so agents can pull customer data and trigger downstream actions.

Use Cases

For high-volume SaaS support: A SaaS company routes routine billing, account, and troubleshooting questions to Decagon’s agents, freeing human agents for complex escalations.
For global consumer brands: A multinational brand supports customers in several languages from a single knowledge base, without hiring region-specific agents.
For regulated enterprises: A company in a sensitive industry deploys Decagon under its SOC 2 / ISO 27001 / GDPR controls to automate support while meeting compliance obligations.

Pricing MODELS

Enterprise

Pricing Summary

Enterprise-only, custom-quoted pricing, generally cited in the roughly $95K–$400K/year range, billed on an outcome (per-resolution) basis rather than per seat. No self-serve or published list pricing.