PublicAI provides a decentralized Web3 platform that connects AI developers with a verified global pool of contributors for multimodal data collection and annotation. The system uses AI‑assisted pre‑labeling, on‑chain validator review, and cryptocurrency payments to deliver high‑accuracy text, audio, video, image, and 3D LIDAR datasets via a scalable API suite.
Funding
Funding not disclosed
Founders
Product
Problem
AI developers need large, high‑quality multimodal datasets for training and evaluating models, but traditional crowdsourcing services impose high costs, inconsistent annotation quality, and cross‑border payment friction.
Solution
PublicAI delivers a decentralized Web3 data infrastructure that links AI builders with a verified global pool of contributors. Clients submit data collection or labeling requests through the PublicAI Marketplace, where the Data Hunter browser extension harvests raw content and AI Assistants generate initial annotations. Human Validators, whose stakes can be slashed for poor performance, review and certify the outputs, ensuring reliable quality via on‑chain reputation and skill‑validation tests. The platform supports text, audio, video, image, semantic‑segmentation, and 3D LIDAR data, and exposes a scalable API suite with vector‑database indexing and real‑time processing. Payments are settled in cryptocurrency, removing banking barriers and enabling unlimited annotator scaling. A free tier provides a baseline annotation quota, while pay‑as‑you‑go usage and an enterprise Pro subscription unlock additional capacity and advanced features.
Target Audience
Primary customers are AI model developers and enterprises that require custom, high‑quality multimodal datasets, as well as labeling service providers seeking a scalable, on‑chain workforce.
Features
- Over 3.5 million on‑chain verified contributors, with skill tests and stake‑slashing for quality assurance.
- AI‑assisted pre‑labeling pipeline followed by validator review, reducing manual effort while maintaining >98 % annotation accuracy.
- Multi‑modal data support: text, audio (24 kHz / 16‑bit WAV), video, image segmentation, and 3D LIDAR point clouds.
- Data Hunter Chrome extension enables seamless, one‑click data collection from social media and GPT conversations.
- PublicAI Data Hub consensus voting rewards contributors with $PUBLIC tokens for dataset validation.
- API suite with vector‑database retrieval, customizable endpoints, and real‑time streaming for enterprise integration.
- End‑to‑end encryption (TLS in transit, AES‑256 at rest) and role‑based access controls for data security.
- Unlimited annotator team size; billing is based solely on completed annotation work.