DecagonUnclaimed

Decagon is a platform for building, testing and running AI agents that talk to a company's customers. The page describes an AI concierge that handles support and service conversations across chat, voice and email, from one shared intelligence layer.

The product is organised around three stages. Teams build an agent by defining its workflows in natural language through what Decagon calls Agent Operating Procedures (AOPs), instead of complex configuration languages. They optimise it with testing, simulations, live A/B experiments and always-on QA. They scale it with an analytics suite that reports on conversations and customer feedback.

Decagon says its channels share one set of rules and memory, so a customer gets consistent answers whether they write in chat, call or send an email. Voice, chat and email are each offered as separate products that can be customised to a brand's tone.

The page is aimed at large enterprises. It names customers in financial services, travel and hospitality, healthcare, technology, retail, telecommunications and insurance, and includes customer quotes about faster setup and lower maintenance work. Product updates such as Voice 3 and personal-agent detection are also listed.

The page does not publish pricing. Access is through a demo request, so the cost of a plan is not stated.