What does Decagon do?

Decagon provides a conversational AI platform for building and operating AI agents that handle customer interactions across chat, voice, and email, execute workflows, retrieve data, take actions, and escalate conversations when required.

How are Decagon agent workflows configured?

Decagon uses Agent Operating Procedures, or AOPs, which allow teams to define custom agent logic and workflows in natural language.

Can Decagon agents perform actions in other systems?

Yes. Decagon integrations allow agents to retrieve data and trigger actions through supported applications, APIs, MCP connections, and custom tool integrations.

Does Decagon maintain customer context?

Yes. Decagon documents memory that maintains customer context across sessions and channels while complementing existing systems of record.

Which channels does Decagon support?

Decagon provides a unified platform for chat, voice, and email. Its chat product supports web, mobile, SMS, WhatsApp, and other messaging experiences.

Can Decagon escalate conversations to human agents?

Yes. Decagon supports live chat escalations, email handoffs, and voice transfers to human agents when escalation is required.

Does Decagon support MCP?

Yes. Decagon documents MCP support for connecting AI agents to data systems and applications for real-time data access and actions.

Can teams test Decagon agents before deployment changes?

Yes. Decagon supports simulated conversations, unit testing, versioning, A/B testing, and step-by-step traceability for evaluating agent updates.

Does Decagon support voice AI agents?

Yes. Decagon Voice supports natural multilingual conversations, automatic language detection and switching, human escalation, outbound calling, guardrails, and support for more than 70 languages.

How much does Decagon cost?

Pricing is not published on the provided official reference pages.

Decagon

Decagon is a conversational AI platform for building customer support agents that execute workflows, take actions, use business data, maintain context, and operate across chat, voice, and email.

Visit Decagon

About Decagon

Decagon is a conversational AI platform for building, optimizing, and scaling AI agents for customer experiences. Its agents can operate across chat, voice, and email while handling customer support workflows, retrieving business data, taking actions, and escalating conversations when human assistance is required.

Agent behavior is configured through Agent Operating Procedures, or AOPs. These allow teams to define agent workflows and business logic in natural language. Decagon also provides guardrails for sensitive operations such as identity verification and refunds, while technical teams can control integrations, versioning, and agent behavior.

Decagon agents can connect with CRMs, help desks, ticketing systems, knowledge bases, call-center platforms, APIs, custom tools, and MCP-compatible systems. The platform also maintains customer context across sessions and channels and provides testing, simulations, versioning, A/B experiments, traceability, observability, and continuous monitoring.

Key Features

Agent Task Execution: Handles customer inquiries end to end, retrieves data, takes actions in connected systems, answers support questions, processes configured workflows, and escalates conversations to human agents when required.
Multi Step Workflows: Agent Operating Procedures let teams define multi step agent logic in natural language for workflows such as troubleshooting, subscription changes, identity verification, refunds, case intake, and other customer processes.
Tools and Actions: Connects to CRMs, help desks, ticketing platforms, knowledge bases, CCaaS systems, APIs, MCP-compatible systems, custom endpoints, and telephony infrastructure to retrieve data and trigger actions.
Triggers and Automation: Customer conversations across chat, email, and voice can initiate agent workflows. Voice agents can also initiate outbound calls for customer updates, support, and sales interactions.
Context and Memory: Maintains customer context across sessions and communication channels, uses connected knowledge bases, and complements systems of record with portable customer context.
Agent Management: Supports natural-language AOP configuration, guardrails, simulated conversations, unit testing, versioning, A/B testing, decision traceability, observability, analytics, diagnostic recommendations, and always-on Watchtower monitoring.

Pricing

Contact Sales for pricing.

Decagon discusses resolution-based pricing, where businesses are charged based on successfully resolved customer interactions rather than traditional per-seat or access-based pricing. The official site does not publish a specific price per resolution.

Pricing last updated: September 26, 2026 at 12:00 AM

Use Cases

Handle inbound customer requests end to end across web chat, SMS, WhatsApp, and other messaging channels while taking actions when needed.
Process subscription changes by helping customers upgrade, downgrade, or pause plans through configured support workflows.
Handle voice-based case intake by gathering customer information, verifying details, executing the relevant workflow, and escalating to a human agent when necessary.
Guide customers through product onboarding and troubleshooting across chat or voice while maintaining context across interactions.

Pros & Cons

Pros:

Agent Operating Procedures allow teams to define and modify multi step customer support workflows in natural language.
Supports pre-built integrations, APIs, MCP connectivity, and custom tool integrations for retrieving data and executing actions across business systems.
Provides cross-channel memory, simulations, unit testing, versioning, A/B testing, decision traceability, observability, and continuous conversation monitoring.

Integrations

Salesforce, Intercom, Zendesk, Zendesk Sunshine, Confluence, Contentful, Kustomer, Amazon Connect, RingCentral, APIs, MCP, custom tool integrations, custom endpoints, SIP trunking