Deccan AI Secures $25 Million Series A to Boost Post‑Training Services
Key Points
- Deccan AI raised $25 million in a Series A round led by A91 Partners.
- The startup provides post‑training data, evaluation, and reinforcement‑learning services for frontier AI models.
- Key customers include Google DeepMind and Snowflake.
- Approximately 125 employees support a contributor network of over one million, primarily based in India.
- Contributor earnings range from $10 to $700 per hour, with top earners making up to $7,000 monthly.
- Revenue run rate is in the double‑digit million‑dollar range, with 80 % from the top five customers.
- Competitors such as Scale AI, Surge AI, Turing, and Mercor also operate in the AI training services market.
Deccan AI, a San Francisco‑based startup that supplies post‑training data and evaluation work for frontier AI models, closed a $25 million all‑equity Series A round led by A91 Partners with participation from Susquehanna International Group and Prosus Ventures. The company leverages a large India‑based contributor network to deliver services such as expert feedback generation, model evaluation, and reinforcement‑learning environments for customers that include Google DeepMind and Snowflake. With about 125 employees and a pool of over one million contributors, Deccan aims to meet the growing demand for high‑quality, time‑critical data that drives reliable AI deployment.
Funding Milestone and Investor Backing
Deccan AI announced the closing of an all‑equity Series A financing round totaling $25 million. The round was led by A91 Partners and included participation from Susquehanna International Group and Prosus Ventures. This capital infusion is intended to expand the company’s post‑training services and scale its operations.
Business Model Focused on Post‑Training Work
While leading AI labs such as OpenAI and Anthropic develop core models in‑house, Deccan AI provides outsourced services that cover the post‑training phase. These services include generating expert feedback, running model evaluations, building reinforcement‑learning environments, and creating data for models that interact with external APIs. The company also offers an evaluation suite called Helix and an operations automation platform for enterprise clients.
Customer Base and Market Position
Deccan AI’s clientele includes high‑profile AI labs and enterprises, notably Google DeepMind and Snowflake. The firm currently works with roughly ten customers and maintains a couple of dozen active projects at any given time. The post‑training market has grown rapidly alongside large language models, with competitors such as Scale AI, Surge AI, Turing, and Mercor also offering data‑labeling and evaluation services.
Workforce and Contributor Network
The startup employs about 125 staff members and operates a large contributor network based primarily in India. More than one million individuals—students, domain experts, and PhDs—are part of the network, with 5,000 to 10,000 active contributors each month. Earnings for contributors range from $10 to $700 per hour, with top performers earning up to $7,000 a month. Deccan’s focus on a single‑country (India) contributor base is intended to simplify quality management compared to competitors that source talent from 100‑plus countries.
Growth and Revenue Outlook
Deccan AI has experienced rapid growth, expanding tenfold over the past year and achieving a double‑digit million‑dollar revenue run rate. Approximately 80 % of its revenue is derived from its top five customers, reflecting the concentrated nature of the frontier AI market. The company positions itself as a “born GenAI” firm, emphasizing higher‑skill post‑training work from its inception rather than traditional computer‑vision labeling tasks.