DeepFine raises ₩10B Series B to turn factory-floor AI into recurring revenue
What's the deal? South Korea's DeepFine has raised a 10 billion won (about $7 million) Series B to build industry-specific AI platforms for logistics, manufacturing, and heavy industry. Hyosung VenturesDealroom has a profile for this one. Try Dealroom →, POSCO Technology InvestmentDealroom has a profile for this one. Try Dealroom →, and LIG D&A-IBKC joined the round. Including this deal, the company has raised roughly 18 billion won to date.
What does DeepFine do? It builds a spatial-intelligence AI platform that combines smart glasses, vision AI, and field data analysis. Workers view drawings and procedures through smart glasses, while employers collect field data to improve quality, productivity, and safety. Its products include the LOGI.FINE logistics tool, the DAO field-support solution, and the DSC 3D-mapping product.
Who are the customers? DeepFine supplies large domestic logistics firms, global manufacturers, and companies in shipbuilding, defence, and maintenance, repair, and operations (MRO). It is now upgrading its products into enterprise software-as-a-service (SaaS) that links to internal systems such as WMS, ERP, and MES.
Why these investors? The backers have footprints in logistics, manufacturing, heavy industry, and defence — DeepFine's core markets. That gives the round a strategic bent, opening doors for technical validation and business partnerships with potential customers, not just capital.
What's the endgame? DeepFine wants to convert one-off, site-by-site projects into repeatable subscription revenue. The plan: standardise common features and system integrations so each new deployment costs less and spreads across a customer's other sites. Funds will go toward upgrading its AI agent technology, expanding SaaS in logistics and MRO, building industry-standard models, preparing for overseas markets, and hiring R&D staff.
What could go wrong? The payoff hinges less on the amount raised than on speed. "This round's success will be determined by how quickly site-specific implementation experience is converted into standard products," the company noted, translated from Korean. If DeepFine cannot lower custom-build costs while meeting each site's security and workflow rules, project wins may not translate into recurring revenue.
In their words: Chief executive officer Kim Hyun-bae said the round shows DeepFine is "recognised as a company that connects industrial digital transformation to real operational results" (translated from Korean). He added that much field work — inspection, equipment checks, safety management — still relies on manual effort the platform aims to support.
The signal: The round sits in the low end of the funding scale by amount, but the investor mix points to where industrial AI is heading: strategic backers using capital to test technology on their own production lines. For AI vendors chasing factory floors, proven deployment records and long-run operating data now matter as much as model performance.
Read more: VentureSquare
Image credit: infomatique
