Most organizations say that they are not fully prepared to use generic AI in a safe and responsible manner, Recently according to a McKinse reportThere is an anxiety clarity – to understand how and why AI makes some decisions. While 40% of respondents see it as a significant risk, only 17% according to the report, are actively addressed.
Seoul based Datumo An AI started as a data labeling company and now wants to help businesses to be safe AI with equipment and data that is capable of testing, monitoring and improving their models without the need for technical expertise. On Monday, the startup raised $ 15.5 million, which increased from investors, including Salesforce Ventures, CB Investment, ACVC partners and SBI Investments, by about $ 28 million.
David Kim, CEO of Detumo and a former AI researcher at Korea’s agency for Defense Development, was disappointed with the nature of data labeling, so he came up with a new idea: a reward-based app that anyone earns and earns money in his free time. Startup validated the idea in a startup competition at Kest (Korea Advanced Institute of Science and Technology). Kim co-established Dutumo in 2018 with five KAIST alumni, east as a selectstar.
Even before making the app completely, Detumo secured thousands of dollars in pre-pronounced sales during the customer search phase of the competition, mostly from KAIST alumni leadership businesses and startups.
In its first year, the startup crossed $ 1 million in revenue and received several major contracts. Today, Startup is counted among its customers among its customers, major Korean companies like Samsung, Samsung SDS, LG Electronics, LG CNS, Hyundai, Navar and Seoul-based telecom giant SK Telecom. Many years ago, however, customers began asking the company to go beyond simple data labeling. The 7 -year -old startup now has more than 300 customers in South Korea and in 2024 revenue has been generated around $ 6 million.
Datumo co-founder Michael Hwang told Techcrunch, “They wanted us to score their AI model output or compare with other outputs.” “When we realized: We were already evaluating the AI model – even without knowing.” Hwang said Dutumo doubled the region and focused the first benchmark Dataset of Korea on the AI Trust and Safety.
“We started in data annotation, then expanded the dataset and evaluation due to the LLM ecosystem maturing,” Kim told Techcrunch.
Techcrunch event
San francisco
,
27-29 October, 2025
Investment in Meta’s recently $ 14.3 billion acquisitions such as invested in Data-Labeling Company Scale AI highlights the importance of this market. Shortly after that deal, AI model manufacturer and meta contestant Openi stopped using Scale AI’s services. The meta deal also indicates that competition for AI training data is intense.
Datumo shared some similarities with companies such as Scale AI in the dataset provision, and arised AI with Galileo and in AI evaluation and monitoring. However, it separates itself through its licensed dataset, especially the data crawled from published books, which the company says that the rich structured human argument provides, but according to CEO Kim, it is very difficult to clean.
Unlike your peers, Detumo also offers a full-deformed assessment platform Datumo EveWhich generates testing data and evaluation to check unsafe, biased or incorrect reactions without automatically requiring manual scripting, Kim said. The signature product is a no-code assessment tool designed for non-developers such as policy, trust and safety and compliance teams.
Asked about attracting investors such as salesforce ventures, Kim reported that the startup first hosted a firecide chat with Andrew NG, the founder of Dapiling.AI at an event in South Korea. After the incident, Kim shared the session on LinkedIn, which attracted the attention of salesforce ventures. After several meetings and zoom calls, investors increased a soft commitment. The entire funding process took about eight months, Hwang said.
The new funding will be used to accelerate R&D efforts, especially in developing automated assessment equipment for enterprise AI, and to score the global Go-to-Market operation in South Korea, Japan and US the Startups, which has 150 employees in Seoul, also attended Silicon Valley in March.
We are always looking to develop, and techcrunch and by providing some insight into our coverage and events in our perspective and response, you can help us! to fill This survey To tell us how we are doing and get a chance to win the award in return!
