Mercor competitor Deccan AI raises $25M, sources experts from India
Summary
Deccan AI, a startup specializing in post-training data and evaluation for AI models, has secured $25 million in an all-equity Series A funding round led by A91 Partners, with participation from Susquehanna International Group and Prosus Ventures. Founded in October 2024, the company provides services such as improving coding and agent capabilities, training systems for API interaction, and generating expert feedback for frontier labs like Google DeepMind and Snowflake. Deccan AI operates with a San Francisco Bay Area headquarters and a large operations team in Hyderabad, India, employing 125 people and leveraging a network of over 1 million contributors, with 5,000-10,000 active monthly. The startup emphasizes quality control by concentrating its workforce in India, addressing the critical need for highly accurate, time-sensitive data in post-training AI development.
Key takeaway
For CTOs and VPs of Engineering evaluating AI model deployment, Deccan AI's $25 million funding highlights the growing necessity and market for specialized post-training data and evaluation services. Your teams should consider outsourcing these complex, quality-critical tasks to dedicated providers to ensure model reliability and performance in production, especially as AI systems evolve into "world models" and demand for rapid, accurate data intensifies.
Key insights
Outsourcing post-training data and evaluation is crucial for reliable AI models, driving demand for specialized services.
Principles
- Quality is paramount in AI post-training.
- Domain-specific data is harder to scale.
- Concentrating operations can enhance quality.
Method
Deccan AI focuses on high-skill post-training work, including expert feedback generation and reinforcement learning, using an India-centric workforce for quality control, and offering evaluation suites and automation platforms.
In practice
- Utilize specialized services for AI post-training.
- Prioritize quality in data generation.
- Consider India for AI training talent.
Topics
- AI Model Training
- Post-Training Data
- AI Evaluation
- Reinforcement Learning
- AI Workforce Development
Best for: CTO, VP of Engineering/Data, Director of AI/ML, Investor, AI Product Manager, AI Operations Specialist
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by AI News & Artificial Intelligence | TechCrunch.