Skan AI raises $63M to fix enterprise AI with real workflow data
Skan AI raised $63 million in Series C funding to build what it calls a 'context graph of work' by observing how employees actually perform jobs across enterprise software. The company is launching two new products, Skan AI Blueprint and Skan AI Agents, alongside its existing intelligence offering. The funding and product announcements address a widespread problem in enterprise AI, where only 8% of companies have AI agents in production and 95% of early implementations require complete redesign.
TL;DR
- Skan AI closed $63 million Series C led by Cathay Innovation and Dell Technologies Capital, bringing total funding to roughly $120 million
- Company launches Skan AI Blueprint and Skan AI Agents to complement existing Skan AI Intelligence platform for discovering, modeling, and automating enterprise workflows
- Gartner research cited by Skan shows only 8% of enterprises have AI agents in production, with 95% of early implementations requiring complete redesign
- Skan's approach uses desktop observation technology to capture how work actually happens across applications, rather than relying on official process documentation and system logs
Why It Matters
Enterprise AI has failed to deliver measurable returns in roughly 95% of pilot programs, according to MIT research. The core problem, Skan argues, is that AI agents are trained on official documentation and system logs that do not reflect how work actually happens in organizations. This gap between documented processes and real workflows is where most enterprise AI initiatives fail.
Business Impact
Companies investing billions in generative AI pilots are seeing dismal results because their AI agents lack accurate understanding of actual business operations. Skan's observation-based approach aims to close this gap by capturing the exceptions, decisions, handoffs, and institutional habits that define real workflows, potentially improving the success rate of enterprise AI deployments.
Key Implications
- Process documentation and system logs are insufficient foundations for enterprise AI agents, suggesting that organizations need new approaches to map actual workflows before deploying automation
- The gap between official procedures and actual work practices represents a significant market opportunity for tools that can observe and model real employee behavior at scale
- Enterprise AI success may depend less on model quality and more on accurate business context, shifting focus from frontier models to workflow intelligence and process discovery
What to Watch
Monitor whether Skan's observation-based approach delivers measurably better results than traditional process mining and documentation-based methods in enterprise deployments. Watch for adoption patterns among large enterprises and whether competitors adopt similar observation technologies. Track whether the success rate of enterprise AI agents improves as organizations implement context-aware approaches.
Subscribe to the newsletter
The latest stories and analysis, delivered to your inbox.
Free. No spam. Unsubscribe any time.
