# Decagon

> Enterprise AI agents that resolve customer conversations across chat, email and voice

- Category: [AI Customer Support](https://finderatlas.com/categories/ai-customer-support)
- Pricing: Paid
- Website: https://decagon.ai
- Platforms: Web
- Made by: Decagon, founded 2023, San Francisco, CA
- Last updated: 2026-10-19
- FinderAtlas page: https://finderatlas.com/tools/decagon

## Overview

Decagon is an enterprise customer service platform that builds, tests and runs AI agents across chat, email and voice. Companies define how the agent should behave in Agent Operating Procedures (AOPs), which describe workflows in natural language, and connect the agent to their existing systems through pre-built connectors. The agents are meant to resolve customer conversations end to end rather than only deflect them to a help article.

## Key features

- Agent Operating Procedures (AOPs)
- Voice, Chat and Email Agents
- Pre-built Integrations
- Experiments
- Testing & QA
- Watchtower
- Insights & Reporting
- Suggestions
- Duet
- Personal Agent Detection

## Pricing

- **Enterprise** (Custom): Quote-based contracts for support organizations

## Best for

- Large companies with high conversation volume that want AI agents handling chat, email and voice with testing and QA built in

## Who uses it

- Customer Support Leaders
- CX Operations Teams
- Enterprise Support Organizations
- Contact Center Managers

## Strengths

- Covers voice, chat and email in one platform rather than chat only.
- Natural-language AOPs let non-engineers define agent workflows.
- Built-in A/B experiments, simulations and always-on QA support ongoing tuning.
- Large enterprise customer list including airlines, telecoms and fintech companies.
- Offers a per-resolution pricing option where escalations are not charged.

## Limitations

- No published pricing; every deal requires a sales conversation.
- No free tier or trial advertised, which limits evaluation by small teams.
- Reviewers have noted limited transparency into why the AI made a given decision.
- No public API or developer documentation was found on the official site.

## Not recommended for

- Small teams or startups needing self-serve signup, published pricing or a free tier

## FinderAtlas editorial scores (out of 10)

- Ease of Use: 7.5/10
- Feature Depth: 8.7/10
- Value for Money: 6.8/10
- Performance & Speed: 8/10
- Documentation & Support: 7.2/10

## Alternatives

- [Assistly](https://finderatlas.com/tools/assistly): AI chat assistant that answers website visitors and captures leads around the clock ([compare](https://finderatlas.com/compare/assistly-vs-decagon))
- [Fin](https://finderatlas.com/tools/fin): AI customer agent that resolves support conversations across chat, email and voice, billed per outcome ([compare](https://finderatlas.com/compare/decagon-vs-fin))
- [Siena AI](https://finderatlas.com/tools/siena-ai): AI customer service agents that resolve e-commerce support tickets in your brand voice ([compare](https://finderatlas.com/compare/decagon-vs-siena-ai))
- [Crescendo](https://finderatlas.com/tools/crescendo): AI-native customer experience platform with AI agents and human specialists on one stack ([compare](https://finderatlas.com/compare/crescendo-vs-decagon))
- [Wonderful](https://finderatlas.com/tools/wonderful): Enterprise AI agents and operating layer for customer service across voice, chat and email ([compare](https://finderatlas.com/compare/decagon-vs-wonderful))
- [Sierra](https://finderatlas.com/tools/sierra): Enterprise AI agents that handle customer conversations across voice, chat and messaging, priced by outcome ([compare](https://finderatlas.com/compare/decagon-vs-sierra))

## Recent updates

- 2026-01-01: Reported $250 million funding round at a $4.5 billion valuation (Bloomberg)

## FAQ

**What channels does Decagon support?**

Decagon's agents handle voice, chat and email, managed through a single platform.

**How does Decagon charge?**

Pricing is not published. Decagon describes per-conversation and per-resolution models, with quotes provided by its sales team.

**How are agents configured?**

Teams write Agent Operating Procedures, which describe workflows in natural language, and connect them to systems through pre-built connectors.

**How can I test an agent before it goes live?**

Decagon provides Testing & QA simulations at scale and live A/B Experiments, plus Watchtower for ongoing quality monitoring.

**Who uses Decagon?**

The site names customers such as Deutsche Telekom, American Airlines, Delta Airlines, Notion, Duolingo, Chime, Rippling, Hertz and Oura.

**Does Decagon offer a free plan or self-serve signup?**

No free plan or self-serve signup is advertised; the site directs visitors to request a demo.