Decagon raises $250M at $4.5B valuation to automate customer support
What’s the deal? AI customer support startup Decagon has raised $250 million in a Series D round, valuing the company at $4.5 billion. The round was led by Accel, with participation from existing and new investors, as Decagon scales software that automates customer service using large language models.
Founded in 2023, Decagon sells AI agents that handle customer queries for large enterprises, positioning itself as a replacement for traditional outsourced call centres. The company says customers use its tools to reduce costs while improving response speed and quality.
Why now? Enterprises are rapidly adopting generative AI to automate labour-intensive functions, and customer support is one of the clearest use cases. Rising wages and pressure to cut operating costs have made automation more urgent.
Decagon has also benefited from strong early traction with large customers, allowing it to raise capital quickly at a steep valuation increase. The Series D comes less than a year after its previous round, reflecting intense investor demand for AI infrastructure plays.
What could go wrong? Competition in AI-powered customer support is heating up, with incumbents, startups, and big tech firms all targeting the same budgets. Switching costs may be lower than expected if customers see these tools as interchangeable.
There are also risks around reliability, data security, and regulatory scrutiny as AI systems handle sensitive customer interactions. Any high-profile failures could slow adoption or trigger tighter controls.
The signal: The deal highlights how quickly AI-native startups can reach multi-billion-dollar valuations when tied to clear enterprise demand. Investors are increasingly backing application-layer AI companies, not just model builders.
It also shows that customer support, long dominated by outsourcing firms, is being reshaped by software-first automation. As generative AI matures, similar disruptions are likely across other back-office functions.
A.M.
