According to marketing documents obtained by TechCrunch and two informed sources, the startup Mercor, which specializes in providing AI model training experts for tech companies such as OpenAI and Meta, is negotiating with investors for its Series C funding round. The company is currently seeking a valuation of $1 billion or higher, an increase from the previously discussed target of $800 million.
According to two sources, the venture capital firm Felicis, which has previously invested in Mercor, is considering increasing its investment in the Series C round. However, Felicis declined to comment on this. A source said that the final terms of the deal may still change.
Mercor told potential investors that the company has received multiple investment offers. According to a previous report by The Information, venture capital firms have been proactively reaching out to Mercor with investment offers valuing the company at up to $1 billion.
TechCrunch learned that the company has already brought in at least two new investors, who are raising funds through special purpose vehicles for this potential deal. The company's previous round of financing was a $100 million Series B round announced in February of this year, with a valuation of $200 million, led by Felicis.
Mercor, founded in 2022, has an annualized revenue of nearly $450 million. In February this year, the company told TechCrunch that its annual revenue had reached $75 million. In March, Mercor's CEO Brendan Fudali posted on X that the company's annual recurring revenue had reached $100 million.
According to one informed source, Mercor told investors that it is expected to reach the milestone of $500 million in annual recurring revenue faster than Anysphere, the developer of the AI programming assistant Cursor. Anysphere reached this goal about a year after launching its product. Unlike Anysphere, which is still losing money, according to Forbes, Mercor achieved a profit of $6 million in the first half of this year.
Mercor's business model involves providing professional domain experts, including scientists, doctors, and lawyers, for enterprises to train AI models, charging fees for intermediary services and matching. The company claims to provide data annotation contractor services for five top AI laboratories, including Amazon, Google, Meta, Microsoft, and OpenAI, as well as Tesla and NVIDIA. According to sources, a significant portion of the company's revenue comes from some of these brands, including OpenAI.
To further diversify its business model, Mercor has been telling investors that the company is adding more software infrastructure for reinforcement learning. Reinforcement learning is a training method that enables models or agents to integrate feedback and improve over time by validating or questioning their decisions. The company also plans to eventually build an AI-driven hiring marketplace platform.
However, Mercor faces challenges from multiple competitors. It is reported that Surge AI is in talks to raise funds at a valuation of $2.5 billion. Additionally, other data annotation companies like Turing Labs and Scale AI are expanding into the reinforcement learning service sector. Some industry insiders believe that OpenAI's recent launch of a recruitment platform could lead the AI giant to create its own human-expert-driven reinforcement learning training service.
When contacted by TechCrunch, Fudali stated that the company is not actively seeking financing and rejects investment offers every month. He also said that the company's annual recurring revenue exceeds $450 million. However, he clarified that the company's revenue includes the total amount paid by customers to Mercor for services, with contractors then receiving their share. He added that this is a common accounting practice recommended by audit firms, and competitors like Surge AI and Scale AI also use this approach.
This startup was co-founded in 2023 by three Thiel Fellows and Harvard dropouts: CEO Brendan Fudali, CTO Adash Shiremat, and COO Surya Mida. All three co-founders are currently under 30 years old. To take the company to the next level, Mercor recently appointed Sundip Jain, who previously served as Chief Product Officer at Uber and has decades of experience, as the company's first President.
Notably, Mercor has recently been sued by its competitor Scale AI, accusing it of stealing trade secrets. Scale AI claimed that a former employee who later joined Mercor stole over 100 confidential documents related to Scale's customer strategies and other proprietary information.
From an industry development perspective, Mercor's rapid growth reflects the explosive demand for specialized talent in AI model training. As large language models and AI applications continue to evolve, the demand for high-quality training data and domain expertise is also continuously increasing.
The company's success also highlights the important value of human resources services in the AI supply chain. Although AI technology itself receives a lot of attention, in reality, the training and optimization of AI models still require a large amount of human involvement, especially from experts in various professional fields.
However, the competitive environment facing Mercor is also intensifying. In addition to traditional data annotation companies, some AI giants are also considering building similar services in-house, which could pose a challenge to third-party service providers like Mercor.
If Mercor successfully completes this $1 billion valuation funding round, it will mark a further increase in the valuation levels of the AI services industry and reflect investors' confidence in the long-term prospects of this niche field. However, in the current market environment, how to maintain profitability and competitive advantage while growing rapidly will be a major challenge for the company.
