Seoul-based Datumo raises $15.5M to take on Scale AI, backed by Salesforce

Most companies say they are not fully ready to use the generator AI in a safe and responsible way, According to the recent McKINCY reportThe An anxiety is interpretation – understanding how and why AI makes specific decisions. Although 40% of the respondents see it as a significant risk, only 17% of the report is actively addressing it.

Seol -based Data An AI data has begun as a labeling company and now wants to help traders to create safe AI with equipment and data that is without technical skills needs – enables their models to test, observe and improve. The startup on Monday collected $ 15.5 million, which brings about $ 28 million to its total, from investors, including sales forms, KB investments, ACVC Partners and SBI investment.

Detumo chief executive David Kim and former AI researcher of the Korean Agency for Defense Defense Defense Defense Defense Defense were disappointed because of the time-consuming nature of the data labeling, so he came up with a new idea: a award-based application that allowed them to label data in extra time and earn money. The startup validated the idea in a startup competition in Caust (Korea Advanced Institute of Science and Technology). Kim co-founded Datomo, formerly known as Selectstar, in 2018, in addition to five Kaiste alumni in 2018.

Before the app was fully built, most of the Caeste-led businesses and startups secured several thousand dollars pre-agreement from the Detumo competition customer invention.

In the first year, startup has surpassed $ 1 million and has achieved several key contracts. Today, startups have calculated the major Korean companies like Samsung, Samsung SDS, LG Electronics, LG CNS, Hyundai, Navar and Seol-based telecom giant SK Telecom. Several years ago, clients began to ask the company to go out of general data labeling. The seven -year -old startup now has more than 300 clients in South Korea and earned about $ 6 million in 2024.

“They wanted we would score their AI model outputs or compare them with their other outputs,” Detumo co-founder Michael Hawang told Techchen. “When we realized: we’re already evaluating the AI model – even don’t know it.” Huang added that Datomo doubled the region and published the Korea’s first benchmark dataset focusing on the AI Trust and Protection.

“We started in data vaccines, then the LLM was matured in the pregnant datasate and evaluation,” Kim told TechCrunch, “Kim told Techchen.

Data Labeling Agency Investing in Meta at AII AII Investing Like Acquisition of $ 14.3 Billion Highlighting the importance of this market. Within a short period of time of this agreement, the AI model manufacturer and meta contestant Open -scale AI has stopped using AI’s servicesThe Meta Deal also indicates that competition for AI training data is intensified.

Detumo Pritrening Datasate Praying Sharing some similarities with companies such as scale AI and share it with it Galileo And AII To evaluate and observe AI. However, it distinguishes itself through its licensed datasets, especially from the published books, the company says that the company provides rich structural humanitarian logic, but it is notoriously difficult to clearly, according to CEO Kim.

Not like its peers, Detumo also provides a full-stack evaluation platform Data assessmentKim added, which produces automatically test data and evaluation to test unsafe, biased or incorrect response without the need for manual scripting. The signature product is a no-code evaluation equipment designed for non-sectors such as policies, beliefs and protection and consent parties.

When asked about attracting investors like the Salesforce Venture, Kim explained that the startup had earlier organized a fireside chat with Andrew NG, the founder of Dippeliating. After the event, Kim shared the session on LinkedIn, which caught the attention of the salesforce initiative. Following several meetings and zoom calls, investors increased a soft promise. The entire funding process took about eight months, Huang said.

The new fund will be used to accelerate the R&D attempt, especially in the development of automatic evaluation equipment for Enterprise AI and to scale the Go-to-Market operations worldwide in South Korea, Japan and the United States, in COL, the COL has established a presence in Silicon Valley in March.

We are always looking for developed and by providing some insights about your outlook and reactions to our coverage and events in our coverage and events, you can help us! Meet This survey We will have the opportunity to win a reward in telling us how we are doing!