Aldagram raises 2 billion yen to build AI workflows into KANNA
The AMW Read
The financing incrementally advances a field-operations software provider's AI strategy, with impact concentrated in its vertical and no demonstrated agent-level outcomes yet.
Aldagram raises 2 billion yen to build AI workflows into KANNA
Aldagram announced a 2 billion yen Series B on September 2, bringing cumulative funding to approximately 5.13 billion yen. JAFCO Group, MonotaRO, and Panasonic Electric Works participated. The company plans to use the financing to develop its KANNA field-work management services into an AI platform and accelerate international expansion. Its strategy has three layers: AI-assisted data capture, connections to external systems and agents, and development of a KANNA agent for workflow assistance and automation.
The AI-market significance lies in enterprise operations: turning an existing workflow application into a surface where AI can access usable records and support work. Aldagram says field information remains scattered across paper, whiteboards, and verbal exchanges, limiting what AI can do. It plans to embed AI OCR and other capture functions to reduce data-entry work, while its KANNA MCP server, introduced in July 2026, lets external AI agents access KANNA data and functions. The competitive question is whether this combination makes field workflows sufficiently accessible and dependable for automation. An integration interface alone does not demonstrate that agents can complete operational tasks reliably.
For builders and investors, the concrete diligence priority is the transition from collecting records to executing useful workflows. KANNA's existing project-management and digital-reporting products provide a starting point, but the proposed agent layer still needs evidence of task completion, accuracy, and reduced manual effort. Adoption of the underlying application should therefore be assessed separately from adoption of its AI capabilities. The funding supports development and expansion; the article does not establish AI-specific revenue or measured productivity gains from the planned agent.