DecideAI

DecideAI focuses on creating reliable, scalable, and affordable AI models by combining open-source technology with a decentralized training and data ecosystem. Its mission is to let users, contributors, and developers collaboratively build, verify, and improve language models, while keeping data transparent and secure on-chain.​

Key features

• Decentralized training via Decide Protocol using Reinforcement Learning from Human Feedback (RLHF) to improve model quality and accuracy.​
• Decide ID for verifying contributors and AI “talent,” helping ensure trustworthy data and high-quality human feedback.​
• Decide Cortex (coming soon) as an access layer for curated datasets and pre-trained models created through the protocol.​
• On-chain AI applications like GPT‑2 and Redactor, fully hosted on blockchain infrastructure for transparency, security, and censorship-resistant operation.​

Use cases

• Teams or startups that want to train specialized LLMs (for niche domains or languages) with transparent, reward-based human feedback.​
• Organizations needing verifiable contributor identity and data quality when sourcing annotations or AI-related work.​
• Developers and researchers who want to experiment with fully on-chain models and datasets, or integrate secure, open-source LLMs into dApps.