A new McKinsey report shows most organizations aren’t fully ready to use generative AI responsibly, with explainability emerging as a top concern. While 40% of executives see it as a significant risk, only 17% are actively addressing it.
Seoul-based Datumo aims to close that gap. Originally launched in 2018 as a reward-based data labeling app, the startup now offers no-code tools for testing, monitoring, and improving AI models, without requiring technical expertise. On Monday, it announced a $15.5 million Series A raise, bringing its total funding to about $28 million. Backers include Salesforce Ventures, KB Investment, ACVC Partners, and SBI Investment.
CEO David Kim, a former AI researcher at Korea’s Agency for Defense Development, co-founded Datumo with five KAIST alumni. After securing early traction, the company evolved beyond annotation to create Korea’s first benchmark dataset focused on AI trust and safety. Today, it serves over 300 clients including Samsung, LG, Hyundai, Naver, and SK Telecom, and reported $6 million in 2024 revenue.
Datumo’s flagship platform, Datumo Eval, automatically generates test data and evaluations to detect unsafe, biased, or incorrect AI outputs. The tool is aimed at non-developers on trust and safety, policy, and compliance teams. The startup’s differentiated datasets, sourced from published books, are designed to capture structured human reasoning.
Related: The Browser Company debuts $20/month Pro plan for its AI browser
With the new funding, Datumo plans to accelerate R&D on automated AI evaluation tools and expand in South Korea, Japan, and the U.S. Its global ambitions also include scaling its recently established Silicon Valley presence.
