Skan AI Raises $63M to Watch How Office Staff Actually Work, Then Build Agents That Copy Them
Skan AI records what staff do on screen and builds agents that replicate their work. Seven of the ten biggest US banks are already customers.
August 12, 2026 – 1:10 pm
Skan AI announced $63 million in funding on Wednesday, co-led by Cathay Innovation and Dell Technologies Capital. The money isn’t buying a model; it’s about understanding how people at large companies actually do their jobs.
Skan builds this understanding by watching them work, using software that takes screenshots of employee desktops. This data is anonymized and abstracted, with only metadata transmitted to the analytics platform. Skan emphasizes that personal messages and passwords are never captured.
The company boasts impressive numbers:
- Revenue grew more than 300% year-on-year.
- Net dollar retention averages 150%.
- They’ve processed over 25 billion work signals.
- A quarter of the Fortune 50 and seven of the ten biggest US banks are customers, along with three major insurers.
Skan highlights a bank example: observing 11.2 million context switches across 1,500 finance professionals led to $37 million in identified operational friction and significant cost and throughput improvements.
The Key Insight: Skan’s value lies not just in observing work but in using those observations to train AI agents that can perform the tasks.
This isn’t a new concept, but Skan emphasizes a unique selling point: while process and task mining have automated workflows based on observation, Skan builds agents that implement those workflows autonomously.
Skan cites Gartner research claiming only 8% of enterprises have agents in production, with 95% of early implementations needing redesigns (access to this report is behind a subscription).
Skan’s CEO, Avinash Misra, frames the offering as a data problem, not a model problem: "You cannot fix a source data problem downstream. Better models will not solve it. Better prompts will not solve it."