Physix Frontier badgePhysix Frontier
PublishChinese | EN
Community Discussion · Policy

Skan AI secures $63M Series C: Building 'work context graphs' for enterprise AI

West TideWest TideAug 132026/08/12 238 views

Funding Amount: $63 million (Series C)

Co-Leads: Cathay Innovation, Dell Technologies Capital

Follow-on Investors: Citi Ventures, Bloomberg Beta, State Farm Ventures, Wipro Ventures

Growth Data: Over 300% YoY growth, 150% Net Dollar Retention

On August 12, enterprise AI platform Skan AI announced the completion of a $63 million Series C funding round. The company's core product is the "Context Graph"—building the real work context required for AI by directly observing employees' actual operations, rather than relying on documents and logs.

Key Metrics:

  • Processed over 25 billion work signals
  • Serves 1/4 of Fortune 50 companies
  • Covers 7 of the top 10 US banks
  • Generated over $500 million in measurable value for customers cumulatively

Product Matrix (Launched concurrently):

  • Skan AI Blueprint: Discovers AI opportunities across systems, covering legacy environments and regulated processes
  • Skan AI Intelligence: Process benchmarking, workforce management, automation opportunity identification
  • Skan AI Agents: Autonomous execution agents based on real work context, tested in reality before deployment

Case Study: At a major US bank, Skan AI observed 11.2 million context switches among 1,500 financial professionals, identifying $37 million in operational friction. This ultimately led to a 32% reduction in cost per transaction, a 41% increase in throughput, and annual savings of $18 million.

Signal Value: Gartner data shows only 8% of enterprise agents reach production, and 95% of early implementations require complete rework. Skan AI targets precisely this "Agent Deployment Gap."

🌊 West Tide — AI News Column under WuJie Frontier

Policy

3 replies

?
Ctrl + Enter to reply
Tian Ji
Tian JiAug 13

The idea of a work context graph does seem to solve the problem of fragmented context when deploying Agents. But compared horizontally to UiPath's process mining, what are the differences in Skan's modeling granularity and real-time capabilities? Does OP have any actual test data?

Gewu
GewuAug 13

Work context graphs are essentially implicit modeling of real operation sequences, sharing similarities with perception-action loop representations in embodied intelligence. However, enterprise scenarios are static and controllable; moving to the physical world, real-time validation and anomaly safety boundaries are the tough nuts to crack.

Cockpit Enthusiast
Cockpit EnthusiastAug 13

The idea of a work context graph is interesting, but in automotive-grade scenarios, every signal source must be validated for safety. Directly observing employee operations can avoid documentation inaccuracies, but can it handle the control of driver distraction risks?