D
Deccan AI generates training data for large language models and reasoning systems, with a stated talent pool of 500,000+ specialists across 25+ domains. The tech stack is frontend-heavy (React, TypeScript, GraphQL) paired with PyTorch and Hugging Face on the backend, suggesting a distributed labeling platform rather than a pure consulting service. Current hiring pressure spans engineering, data, and research roles at an accelerating pace, while internal pain points reveal operational friction: same-day turnaround demands, rapid scaling of supply functions, and gaps in evaluation metrics—classic signals of a data services business managing quality and velocity simultaneously.
Tech Stack 33 technologies
Core StackReact JavaScript Redux GraphQL TypeScript Webpack GitHub Bitbucket Python PyTorch Hugging Face FastAPI Docker Snowflake Adobe Photoshop Adobe Illustrator Adobe InDesign Canva Figma HTML5 CSS MobX Babel Git Transformers BERT Google DeepMind Google Workspace Excel Google Sheets+3 more
What Deccan AI Is Building
◆Challenges
- Same day turnaround for complex data needs
- Engagement and retention
- Improving employee experience at scale
- Consistency across teams
- Rapid hiring needs
- Multiple requisition management
- Scaling supply function
- Existing benchmarks insufficient
- Gaps in evaluation metrics
▲Active Projects
- Benchmark design and evaluation dataset creation
- Multi-agent ai systems
- Rag pipelines
- Nlp chatbots
- Employee listening programs
- Onboarding programs
- Culture and belonging initiatives
- Business strategy
- Fundraising
- Product launches
Hiring Activity
15 roles · 8 in 30d
Department
Seniority
Company intelligence
