Perceptron
Category: AI Infrastructure
Perceptron is a decentralized AI data network that crowdsources real-world training data through a global mesh of user nodes, enabling AI companies to source, verify, and monetize high-quality datasets at reduced cost. Perceptron was founded in 2024. The company is led by Peter Anthony. Based in Dubai, United Arab Emirates. Team size: 11-50. Total funding raised: $6.5M. Latest round: Strategic Investment. Key investors include Sigma Capital, Selini Capital, QCP Capital, P2 Ventures, CoinDCX Ventures, Momentum6, DeFi Capital, Walrus Foundation, Aethir, Colosseum, GuruDev Capital, Tempo Finance.
- Founded
- 2024
- Headquarters
- Dubai, United Arab Emirates
- Team size
- 11-50
- Total funding
- $6.5M
Value proposition
Decentralized AI data mesh that provides 90% cost reduction vs centralized data providers, real-time geographically diverse data from 800K+ nodes across 150+ countries, and a reward-driven ecosystem where contributors own their data and earnings.
Products and solutions
Perceptron Nodes (browser extension/Android app sharing idle bandwidth), Perceptron Agents (embedded in Discord, Telegram, WeChat for data labeling/annotation), Data-Questing Platform (commissioning custom datasets from community), $10M AI Data Fund (supporting independent AI developers with up to 5 TB of free data and 5 weeks of infrastructure support), PERC Token (reward and governance token on Solana)
Unique value
World's first decentralized AI data mesh — turning idle consumer bandwidth into a globally distributed vantage point on the live web, enabling AI companies to source verified, real-world datasets in days at 90% lower cost than centralized alternatives.
Target customer
AI companies and startups needing affordable training data; hedge funds and quant traders needing real-time market intelligence; e-commerce businesses needing cross-border pricing/ranking data; independent AI developers
Industries served
AI/ML Training, Financial Services/Quant Trading, E-commerce, Crypto/Web3
Technology advantage
Dual-engine architecture (Nodes for data gathering + Agents for edge labeling); blockchain-based trust layer (PERC tokens/NFTs for reputation and rewards); 800K+ node network spanning 150+ countries; 90% cost advantage vs centralized providers; Solana-based tokenomics
How they differentiate
Decentralized node network (800K+ nodes) providing geographically diverse, real-time data at 90% lower cost than centralized providers; contributors own their data and earnings; human-in-the-loop annotation via data-questing platform; merged with BlockMesh (700K nodes) in mid-2025 for combined scale
Main competitors
Grass (decentralized data scraping network), Masa (decentralized data marketplace for AI), Nuklai (decentralized data infrastructure)
Key partnerships
BlockMesh (merged June 2025 — 700K+ node network), Aethir (decentralized compute infrastructure), Walrus Foundation
Notable customers
Everlyn AI (text-to-image/video platform)
Major milestones
Mid-2025: Merged with BlockMesh (700K+ nodes), June 2025: Launched $10M AI Data Fund, July 2026: Raised $6.5M strategic round, 2026: Scaled to 807K+ nodes with 300K+ daily active users, 2026: Launched data-questing platform
Growth metrics
807K+ nodes onboarded; 300K+ daily active users; 150+ countries covered; 90% lower operational costs vs centralized providers
Market positioning
Emerging leader in the DeAI (Decentralized AI) data stack, positioned between centralized scraping tools and closed data partnerships — targeting the growing demand for affordable, diverse training data as centralized data sources become increasingly paywalled and depleted.
Geographic focus
Global (150+ countries with node network); HQ in Dubai, UAE
About Peter Anthony
Crypto native since 2019; Founder of The House of Crypto (YouTube channel with ~300K followers); Background in forex, stocks, and traditional markets trading; University of Southampton
Latest news about Perceptron
More AI Infrastructure companies
Official website: https://perceptrons.xyz
