Decagon is an AI concierge platform that automates customer service across voice, chat, and email. Read about its features, pricing model, and best-fit users.
Decagon is an AI customer service company that builds what it calls an AI concierge: agents that handle support conversations over voice, chat, and email on behalf of a business. Rather than a single chatbot widget, Decagon's platform is built for enterprises that need agents woven into existing support operations, with memory that carries context across channels so a customer doesn't have to repeat themselves when moving from chat to a phone call.
Founded in 2023 by Jesse Zhang and Ashwin Sreenivas, Decagon has grown quickly and raised over $480 million in funding, reaching a reported $4.5 billion valuation by early 2026. The company has positioned itself as one of the more prominent players in applying large language models to customer support at scale, with named customers spanning neobanks, travel companies, and consumer subscription apps.
Decagon's core feature is what it calls Agent Operating Procedures, a natural-language way to define how an agent should behave in different scenarios without writing traditional workflow code. This is paired with simulation and A/B testing tools so support teams can test changes to an agent's behavior before pushing them live, rather than relying purely on post-launch monitoring.
The platform also includes an analytics and observability layer that surfaces customer insights, deflection rates, and quality assurance scoring, plus a library of integrations meant to connect the agent to a company's existing helpdesk, CRM, and internal tools. Decagon reports outcomes such as 70-80% resolution rates and significant cost reductions for some customers, though these figures come from Decagon's own case studies rather than independent audits.
Decagon does not publish list pricing on its website; the homepage leads directly to a demo request rather than a pricing page or self-serve signup. Based on industry reporting, the company typically prices around two models: a per-conversation fee charged for every interaction the AI touches, or a higher per-resolution fee charged only when the AI successfully resolves the issue.
Third-party marketplace data suggests enterprise contracts commonly range from roughly $100,000 to over $900,000 per year, with a reported median near $400,000 annually. Because pricing is individually negotiated, actual costs depend heavily on conversation volume, channel mix, and the complexity of the integrations required.
Decagon does not publish pricing. It typically charges enterprise customers per conversation or per resolution, with contracts reportedly ranging from roughly $100,000 to over $900,000 annually depending on volume.
No. Decagon's website routes prospective customers to a demo request rather than a self-serve signup, and pricing is negotiated directly with its sales team.
Decagon agents operate across voice, chat, and email, sharing memory so context carries over when a customer moves between channels.
Publicly named customers include Chime, Duolingo, ClassPass, and Rippling, spanning fintech, consumer subscriptions, and workforce software.
Decagon markets its Agent Operating Procedures as a natural-language alternative to code-based workflows, aimed at support and operations teams rather than engineers.