IrisGo, a startup backed by AI pioneer Andrew Ng, is developing a desktop AI assistant that automatically learns and replicates user workflows by observing desktop activity. Rather than requiring explicit programming or setup, Iris watches what users do on their computers and builds models of their recurring tasks, then offers to automate them.

The approach prioritizes usability over technical sophistication—users don't need to write scripts or configuration files. Instead, the system continuously learns from observation and adapts to individual work patterns, making automation accessible to non-technical professionals.

What This Means for Your Business

If Iris delivers on its promise, it could significantly reduce the time professionals spend on repetitive desktop tasks, particularly in roles involving data entry, document processing, or cross-application workflows. Organizations should track this product's maturity, as widespread adoption could meaningfully improve knowledge worker productivity. However, IT teams should also prepare for security and data governance challenges around an AI system that monitors desktop activity and requires careful permission boundaries.