Perle Labs is a decentralized artificial intelligence (AI) data infrastructure platform designed to create expert-validated, human-verified, and on-chain auditable training data. The project describes its infrastructure as a "sovereign intelligence layer for AI," aimed at serving enterprises and government entities that require high levels of trust, transparency, and auditability in their data pipelines. Founded by veterans from the AI data company Scale AI, Perle Labs utilizes the Solana blockchain to provide a transparent and verifiable system for data annotation and model alignment, positioning itself as an alternative to opaque, "black-box" data solutions. ^{[1] [2]}

Overview

Perle Labs was created to address critical flaws in the traditional AI data labeling market, which can compromise the quality and safety of AI models. The core problem it targets is the "garbage in, garbage out" principle, where low-quality, biased, or untraceable training data leads to unreliable and unsafe AI. The platform also aims to mitigate risks such as data poisoning, where malicious data is used to corrupt AI models, and "model collapse," a phenomenon where AI models degrade after being trained excessively on synthetic, AI-generated data. ^{[3] [4]}​

The platform's solution is built on two pillars: verified human expertise and Web3 infrastructure. It moves away from anonymous crowd workers by onboarding and vetting subject-matter experts from fields like medicine, law, and engineering. Every contribution, review, and reward is logged on the Solana blockchain, creating an immutable audit trail for data provenance. This on-chain system provides a verifiable record of an expert's work, allowing them to build a portable, "sovereign" reputation. Perle's vision is to foster an "Expert Economy" that rewards quality and verifiable expertise, contrasting with the traditional "Gig Economy" model that often prioritizes speed and volume. ^{[5] [6]}​

Perle operates a dual-platform model. Perle AI is its enterprise-facing wing, providing data annotation workflows and quality assurance services to AI development teams. Perle Labs is the decentralized, Web3 contributor-facing platform where individuals perform data labeling tasks, earn on-chain rewards, and build their on-chain reputation. ^[2]​

History and Funding

The founding team of Perle Labs is composed of anonymous veterans with experience from the AI data company Scale AI, as well as from organizations such as MIT, Meta, and Amazon. The project's CEO is Ahmed Rashad. The founders' identities are kept private, with the project's public focus centered on the team's collective industry experience. ^{[2] [5]}​

The project announced it had raised $17.5 million in funding. The round was co-led by Framework Ventures and CoinFund, with participation from Protagonist, HashKey Capital, and Peer VC. ^{[4] [1]}​

There are conflicting reports regarding the platform's public launch. One source states that Perle Labs publicly launched "Season 1" of its data labeling platform on November 2, 2023, the same day it announced its funding round. This initial phase was focused on generating preference datasets for Reinforcement Learning from Human Feedback (RLHF). ^[5] However, a separate report published in early 2026 states that the platform officially opened to the public with its "Season 1" launch on January 16, 2026. This launch followed a two-month closed beta period during which testers completed 1.7 million tasks and earned 333 million points. ^[2]​

Technology and Architecture

Perle's technology is designed to transform human expertise into a programmable and portable asset within the AI supply chain. Its infrastructure is built to be auditable and transparent, providing a high-integrity data source for critical AI applications. ^[6]

The Sovereign Intelligence Layer

The core concept behind Perle is the "Sovereign Intelligence Layer," a decentralized network where data is not controlled by a single entity. It allows organizations to access a transparent marketplace for data services, with all activities recorded on-chain. This provides verifiable data provenance, ensuring that high-stakes AI decisions can be traced back to accountable human judgment. The system is designed to provide clients with a 30% improvement in AI model accuracy and a 20% reduction in data annotation errors. Perle claims its data indexing infrastructure is 10 times faster than traditional methods, with 99.9% accuracy in data curation. ^{[1] [5]}

Architectural Layers

The protocol utilizes a modular, four-layer architecture to manage its data and contributor ecosystem: ^[6]

• Layer 1: Data and Task Layer: This layer is responsible for ingesting, processing, and validating various data types, including text, images, video, audio, and sensor streams for AI training.
• Layer 2: Coordination and Reputation Layer: This layer manages the routing of tasks to qualified experts, calculates contributor reputation scores, allocates rewards, and oversees the quality control process.
• Layer 3: Settlement and Record Layer: Operating on the Solana blockchain, this layer immutably records all transactions, task completions, reputation updates, and token transfers, serving as the system's ledger.
• Layer 4: Application and Interface Layer: This layer provides the user-facing components, including dashboards, APIs, and SDKs that allow expert contributors and clients (task requesters) to interact with the platform.

On-Chain Reputation System

A central feature of Perle's architecture is its on-chain reputation system. Instead of rewarding speed, the system tracks contributor performance based on accuracy, consistency, and reviewer feedback. This performance data is used to calculate a multi-dimensional score that contributes to a persistent, on-chain reputation. This public and verifiable track record functions as a digital résumé, allowing contributors to showcase their expertise. A higher reputation score unlocks access to more complex, higher-paying tasks, creating a meritocratic environment where quality work is incentivized. ^[3]

Web3 Infrastructure

Perle is built on the Solana blockchain to leverage its high throughput and low transaction fees. This enables several key functions: ^[3]

• Data Provenance: Every contribution and review is logged on-chain, creating a transparent and immutable audit trail.
• Automated Rewards: Smart contracts are used to automate the distribution of rewards to contributors based on verified task completion and quality scores, ensuring fair and efficient payouts.
• Global Accessibility: The use of Solana facilitates near-instant, low-cost micro-transactions, making it feasible to manage a global network of contributors.

Ecosystem and Business Model

The Perle ecosystem is structured as a three-sided marketplace designed to connect data requesters with a network of verified experts. ^[7]

Participants

• Task Requesters: This group includes AI development teams, frontier AI labs, and companies in regulated sectors like healthcare and autonomous vehicles. They use the platform to source high-quality, reliable data for training, testing, and fine-tuning their AI models. They typically pay for these services in stablecoins. ^{[6] [5]}
• Contributors: This is the network of over 50,000 individuals who perform data work. ^[1] This group includes:
  • Subject-Matter Experts (SMEs): Verified professionals such as doctors, lawyers, and engineers who apply their domain knowledge to specialized tasks.
  • Annotators and Reviewers: Individuals vetted for their skills in data labeling and quality control.
  • Validators: A subset of experienced, high-reputation experts who review the work submitted by other experts to ensure quality and accuracy. ^[3]

Business Model

Perle Labs generates revenue through several channels: ^[6]

• Enterprise Contracts: Large-scale, long-term data programs for corporate and government clients.
• Usage-Based Fees: A self-service model where clients pay based on their platform usage.
• Premium Services: Specialized offerings such as model safety evaluations, design of Reinforcement Learning from Human Feedback (RLHF) programs, and the creation of custom data benchmarks.

A core tenet of the platform is to provide trustable rewards for contributors, with payouts guaranteed and managed through secure smart contract escrows. ^[1]

PRL Token

The PRL token is the native utility and incentive token of the Perle Labs ecosystem, operating on the Solana blockchain. Its primary purpose is to coordinate incentives among all network participants and secure the network. ^[4]

• Token Ticker: PRL
• Blockchain: Solana
• Total Supply: 1,000,000,000 PRL (fixed)
• Contract Address: PERLEQKUNUp1dgFZ8EvyXHdN9d6ZQqfGxALDvfs6pDs

The token allocation is divided among key areas of the ecosystem, with a total of 17.5% of the supply circulating on Day 1 from the Community and Ecosystem funds. The allocations for the team and investors are subject to a 12-month cliff for 25% of their tokens, followed by a 36-month linear vesting period for the remaining 75%. The allocations are as follows: ^[6]

• Community: 37.50%
• Investors: 27.66%
• Ecosystem: 17.84%
• Team: 17.00%

Token Utility

The PRL token is integrated into the platform's core functions: ^{[4] [5]}

• Staking: Experts are required to stake PRL to participate in the network. This stake acts as a security deposit that can be "slashed" (forfeited) for low-quality or malicious work, ensuring alignment with network quality standards.
• Access and Priority: Holding or staking PRL can grant users priority access to premium data-labeling tasks, early access to new product features, and preferential pricing for enterprise clients.
• Rewards: Experts and validators are compensated for their work with rewards paid in PRL.
• Payment Mechanics: While clients pay for services in stablecoins, a portion of protocol fees may be used to programmatically buy back and burn PRL, creating deflationary pressure on the token supply.
• Governance: PRL holders are intended to have the ability to participate in the decentralized governance of the protocol, voting on network parameters and upgrades.

One of the project's guiding principles is a creator-first approach, as stated in its communications:

"Perle is a creator-first platform. We believe that human intelligence, creativity, and expertise are the most valuable resources for building the next generation of AI." ^[3]

Leadership and Governance

Perle is led by CEO Ahmed Rashad, and the founding team includes veterans from Scale AI, MIT, Meta, and Amazon. ^[2]

The project's governance is structured to separate the technology development from the ecosystem stewardship. Perle Labs, Inc. is the core technology company responsible for building the platform. The Perle Foundation, a separate non-profit entity, is responsible for overseeing the PRL token, including its distribution, treasury management, and the future stewardship of protocol governance. ^[4]

This separation is intended to facilitate a transition toward a more decentralized governance model where the community of PRL holders can participate in key decisions. The project's mission is summarized in its messaging:

"We're trying to prove that you can bring greater transparency to the supply chain of data that will train the next generation of AI models... You should know where your data comes from. You should know who worked on it. You should be able to audit this on-chain." ^[5]

Security

Perle Labs has undergone a security audit conducted by the blockchain security firm Halborn. The audit report is available for verification and covers the platform's smart contracts and the PRL token to ensure system integrity. ^{[1] [3]}

Market Adoption and Recognition

Since its launch, Perle Labs has begun onboarding enterprise clients, particularly in specialized fields requiring high data accuracy, such as healthcare and robotics. The company has reportedly started to replace incumbent data-labeling vendors on some enterprise AI development projects. ^[2]

The project has also received industry recognition. Perle was cited in an IEEE Spectrum report on the data labeling industry's role in the development of agentic AI. The report mentioned Perle alongside major industry players like Scale AI and SuperAnnotate, indicating its growing relevance in the field. ^[2]