Deccan AI raises $25M to scale India-based AI training workforce

Deccan AI, a startup headquartered in the San Francisco Bay Area with a large operations team in Hyderabad, has raised $25 million in its first major funding round. The Series A round was led by A91 Partners, with participation from Susquehanna International Group and Prosus Ventures. The company is using the capital to scale its operations, which focus on post-training data and evaluation services for frontier AI labs—companies building core artificial intelligence models. Post-training refers to the work done after core model development, including data generation, evaluation, and reinforcement learning, to make systems reliable for real-world use.

Founded in October 2024, Deccan AI differentiates itself by concentrating its workforce primarily in India to maintain quality. This contrasts with competitors such as Turing and Mercor, which also source contractors from India but operate across a broader set of emerging markets. According to founder Rukesh Reddy, “If you have operations in just one country, it becomes far easier to maintain quality.” The startup’s operational footprint and contributor network include:

• Internal team: The company employs about 125 people.
• Contributor network: A network of more than 1 million contributors, including students, domain experts, and PhDs.
• Active monthly contributors: Around 5,000 to 10,000 contributors are active in a typical month.
• Advanced credentials: About 10% of Deccan’s contributor base holds advanced degrees, such as master’s and PhDs.
• Platform earnings: Earnings on the platform range from about $10 to $700 per hour, with top contributors earning up to $7,000 a month.

The sector has faced criticism over working conditions and pay, with large pools of gig workers often used to generate training data. Despite these industry-wide challenges, Deccan AI reports growth. According to Reddy, the company grew 10x over the past year. However, its revenue is highly concentrated, with about 80% of its revenue coming from its top five customers, which include Google DeepMind and Snowflake. This concentration reflects the narrow nature of the frontier AI market, where core model development remains concentrated among a handful of U.S. companies and a few players in China. Other competitors in the broader data labeling and evaluation market include Scale AI and Surge AI.

Why it matters

Deccan AI is emerging as a key player in the post-training data and evaluation market, which is increasingly being outsourced by frontier AI labs to ensure system reliability.