Sapien

What Sapien does

Sapien is a decentralized data foundry that transforms collective human knowledge into enterprise-grade AI training data. They offer Proof of Quality (PoQ), an open infrastructure for verifiable quality signals in AI pipelines. PoQ allows users to define quality, submit data for evaluation, labeling, or improvement, and receive attestations on-chain. This system uses contributors and validators who stake, work, and reach consensus, with results written on-chain to provide immutable provenance and neutral settlement. PoQ aims to build a trust layer for AI by addressing the quality problem in AI data, where different teams define quality differently and work happens in unauditable tools. It ensures that who created and validated a piece of data, and what consensus was reached, is verifiable.

Products

• Proof of Quality (PoQ)An open infrastructure for verifiable quality signals in AI, adding a trust layer to any AI pipeline. It allows users to define quality, submit data, and receive on-chain attestations of consensus-based validation.

Key capabilities

• Verifiable quality signals for AI
• Decentralized data foundry
• Define quality with Task Definition Spec (TDS)
• Submit data points for evaluation, labeling, or improvement
• Task registered on-chain for contributors
• Contributors stake and submit work
• Validators review via commit-reveal
• Stake-weighted consensus determines acceptance
• Attestation written on-chain
• Immutable provenance
• Neutral settlement
• Data never leaves user systems (only attestation goes on-chain)
• Economic consensus for quality verification
• Drop-in SDKs (JS and Python)
• Every output attested on-chain (who created, who validated, consensus)
• Validators stake real capital; bad work gets slashed
• Full audit trail on-chain
• Adapters for existing tools
• Alpha testnet available
• 1M+ contributors

Use cases

• Training Data validation
• Fine-Tuning data validation
• Evaluation (add verifiable quality gates to eval pipeline)
• Production (real-time verification for safety-critical decisions)
• Agent Workflows (approval gates with full provenance)
• Safety-Critical Systems (real-time verification for robotics, AVs, and high-stakes decisions)
• Labeling & annotation (show training data was checked)
• AI & agent evaluation (grade AI answers with a group, not one person)
• Expert & audit review (turn expert findings into proof)
• Internal QC (replace random spot checks)

AI approach

Sapien provides an open infrastructure for verifiable quality signals in AI, called Proof of Quality (PoQ). It allows users to define quality for their AI tasks, submit data for evaluation, and receive on-chain attestations of consensus-based validation. This system is designed to add a trust layer to AI pipelines, ensuring data quality for training, fine-tuning, evaluation, and real-time verification for production and safety-critical decisions. The system uses economic consensus where contributors and validators stake, work, and reach consensus, with bad work being slashed.

Tech named: computer vision, LLMs, AI models, web3, onchain, ML workflows, smart contracts, protocol design

What it says sets it apart

• Decentralized data foundry
• Verifiable quality signals for AI
• On-chain attestations for trust and provenance
• Economic consensus for quality verification
• Data sovereignty (data never leaves user systems)
• Open infrastructure
• Neutral settlement
• Addresses the AI industry's quality problem by providing auditable proof
• Multidisciplinary team with deep crypto infrastructure and AI data expertise

Funding rounds we track

From the AI funding tracker — rounds as reported by the linked publications.