The platform classifies each AI task, identifies how the tools are being used and applies real-time controls, flagging risks such as staff using personal accounts or exposing sensitive data. Under the hood, Harmonic uses small language models, a compact type of AI model, to read each AI interaction in context so organisations can categorise tasks, spot emerging use cases and set policies. The same engine powers its Usage Intelligence layer, which shows which AI use cases are actually creating value for employees.

The round was led by Crosspoint Capital Partners, with additional support from existing investors, and the startup plans to expand its engineering and go-to-market teams in the US and UK. Paterson claims the custom small language models give the company unique insight into enterprise AI use, saying most vendors struggle to answer boards' risk questions while Harmonic addresses both risk and value.