News
Apr 22, 2026
News
Startups
Artificial Intelligence
Americas
NewDecoded
3 min read

Image by deccan.ai
Deccan AI has secured $25 million in a Series A funding round led by A91 Partners to advance its work in high-accuracy artificial intelligence. The investment marks the first AI-focused move for A91 Partners and includes participation from Susquehanna International Group and Prosus Ventures. This capital injection follows a year of rapid growth for the Mountain View based startup, which has increased its revenue tenfold since its launch.
The company operates as a critical infrastructure provider for the post-training phase of AI, offering pristine data and rigorous evaluation tools. By focusing on the philosophy that accuracy is intelligence, Deccan AI helps frontier labs like Google DeepMind and Snowflake refine their models. Their platform includes the STARK RL environment for reinforcement learning and the Helix suite for monitoring agentic performance in production.
Founded by former Google engineer Rukesh Reddy, the firm leverages a global network of over one million contributors to solve the data bottleneck. This workforce includes a specialized tier of domain experts who handle complex tasks in coding, finance, and STEM. These professionals ensure that the training data used for fine-tuning models remains high-quality and free from the errors common in automated labeling.
Looking forward, the startup plans to expand its EnterpriseOS solution to help Fortune 500 companies deploy AI agents within their existing infrastructure. These bespoke tools are designed to automate back-office workflows while keeping human operators in control of the process. The expansion will be supported by a growing team across the United States and India, including a new office in Bengaluru.
The successful funding of Deccan AI signals a major pivot in the industry from raw model scaling toward the pursuit of reliable, production-ready accuracy. As the market for foundational models matures, the focus is shifting to the human-in-the-loop validation necessary for safe enterprise deployment. This investment reinforces the value of specialized infrastructure that can bridge the gap between experimental AI prototypes and robust execution in the real world.
Related Articles