As the race to build reliable artificial intelligence shifts from raw scale to precision, Deccan AI has raised $25 million in a Series A funding round to expand its post-training and evaluation services. The all-equity round was led by A91 Partners, with participation from Susquehanna International Group and Prosus Ventures.
Founded in late 2024, Deccan AI focuses on the critical “last mile” of AI development. While frontier labs like OpenAI and Anthropic build core architectures, Deccan provides the expert feedback and reinforcement learning necessary to make these models functional in the real world.
The Shift Toward Expert-Driven Training
The industry is moving beyond simple data labeling toward high-stakes evaluation. Deccan’s workforce helps models master complex coding, interact with software APIs, and transition into “world models” that understand physical environments for robotics and vision systems.
Unlike traditional data firms that pivoted from computer vision, Deccan was “born GenAI,” focusing exclusively on high-skill tasks. Its product suite includes Helix, an evaluation platform, and an operations automation suite designed for enterprise clients.
Leveraging India’s Intellectual Capital
While competitors like Mercor, Turing, Scale AI, and Surge AI often source talent from over 100 countries, Deccan has concentrated its operations in India. Founder Rukesh Reddy argues that this focus allows for tighter quality control—a necessity when the tolerance for error in production-ready AI is near zero.
Key Stats and Reach:
- Network: Over 1 million contributors, including PhDs and domain experts.
- Active Engagement: 5,000 to 10,000 contributors active monthly.
- Compensation: Rates range from $10 to $700 per hour for top-tier expertise.
- Clients: High-profile partnerships with Google DeepMind and Snowflake.
Rapid Growth and Market Position
Headquartered in the San Francisco Bay Area with a massive operations hub in Hyderabad, Deccan has seen 10x growth over the past year. The startup currently operates at a double-digit million-dollar revenue run rate, with the majority of its business coming from five major frontier AI labs. By prioritizing depth of expertise over geographic breadth, Deccan is positioning itself as the go-to partner for labs that require high-quality, domain-specific data on extremely tight timelines.
