# As a condition of accessing this website, you agree to abide by the following # content signals: # (a) If a Content-Signal = yes, you may collect content for the corresponding # use. # (b) If a Content-Signal = no, you may not collect content for the # corresponding use. # (c) If the website operator does not include a Content-Signal for a # corresponding use, the website operator neither grants nor restricts # permission via Content-Signal with respect to the corresponding use. # The content signals and their meanings are: # search: building a search index and providing search results (e.g., returning # hyperlinks and short excerpts from your website's contents). Search does not # include providing AI-generated search summaries. # ai-input: inputting content into one or more AI models (e.g., retrieval # augmented generation, grounding, or other real-time taking of content for # generative AI search answers). # ai-train: training or fine-tuning AI models. # use: how AI systems may consume the content (immediate, reference, or full). # ANY RESTRICTIONS EXPRESSED VIA CONTENT SIGNALS ARE EXPRESS RESERVATIONS OF # RIGHTS UNDER ARTICLE 4 OF THE EUROPEAN UNION DIRECTIVE 2019/790 ON COPYRIGHT # AND RELATED RIGHTS IN THE DIGITAL SINGLE MARKET. # BEGIN Cloudflare Managed content User-agent: * Content-Signal: search=yes,ai-train=no,use=reference Allow: / # Training crawlers User-agent: Amazonbot Disallow: / User-agent: Applebot-Extended Disallow: / User-agent: Bytespider Disallow: / User-agent: CCBot Disallow: / User-agent: ClaudeBot Disallow: / User-agent: Diffbot Disallow: / User-agent: Google-Extended Disallow: / User-agent: GPTBot Disallow: / User-agent: omgili Disallow: / User-agent: anthropic-ai Disallow: / User-agent: Claude-Web Disallow: / User-agent: cohere-ai Disallow: / User-agent: MistralAI-Training Disallow: / User-agent: meta-externalagent Disallow: / User-agent: GoogleOther Disallow: / User-agent: Baiduspider Disallow: / User-agent: PetalBot Disallow: / User-agent: AwarioSmartBot User-agent: AwarioRssBot Disallow: / User-agent: Google-CloudVertexBot Disallow: / User-agent: QualifiedBot Disallow: / User-agent: Cotoyogi Disallow: / User-agent: ICC-Crawler Disallow: / User-agent: atlassian-bot Disallow: / User-agent: FishBot Disallow: / User-agent: BorderxBot Disallow: / User-agent: NavuBot Disallow: / User-agent: SemrushBot-SWA Disallow: / User-agent: WARDBot Disallow: / User-agent: magpie-crawler Disallow: / User-agent: KimiBot Disallow: / User-agent: CitibotSiteCrawler Disallow: / # END Cloudflare Managed Content Perle AI
Perle

Perle Is How
Expertise
Gets Into AI.

From Data To Deployment, Perle Helps Teams Hire Smarter With AI-Powered Interviews, Evaluations, And Insights.

SOC 2 Type II Certified

See how frontier models handle code-switching speech, Arabic dialects, Saudi legal documents, sound classification, hard compositional image prompts, multi-agent image loops, and calibrated pre-labeling — or ask us to evaluate yours.

Explore Benchmarks

The Stack

3 Products. 1 Human Standard.

Everything we build serves a single goal: get the world’s best minds into your training loop without losing the texture of how they think.

Programs That Run From Brief To Delivery

Programs That Run From Brief To Delivery

Bring structure to your pipeline with training data, evaluation, and labeling handled end-to-end by expert teams and a unified platform.

Experts Who Raise The Floor On Your Team

Experts Who Raise The Floor On Your Team

Scale your team with vetted specialists using interviews, domain rubrics, and quality signals that maintain high judgment standards.

Embodied Data That Keeps Pace With Your Roadmap

Embodied Data That Keeps Pace With Your Roadmap

Access continuously refreshed, expert-labeled multimodal data designed for teams building real-world robotics systems.

What we work on

The problems frontier teams keep bringing us

When the easy data is exhausted and a model still gets the hard things wrong, this is where we come in.

Training robots to operate in the real world

WhisperMind captures the way real hands move through real spaces, picking up a wet glass, opening a stuck drawer, finding the salt in a stranger's cupboard. Multi-sensor RGB-D, IMU, and force-torque streams labeled by people who've cooked the meal.

Learn more

Stress-testing models a board exam wouldn't catch

Our evaluation network spans 10,000+ licensed physicians, 800 attorneys, and specialists across two dozen disciplines. They write the rubric, run the cases, and flag the failures that benchmarks paper over.

Learn more

Drawing the line a model shouldn't cross

Red-teaming, refusal calibration, and bias auditing led by domain experts, not by people who learned the topic from the prompt. We tell you exactly where your model breaks, in language your safety team can act on.

Learn more
Evaluation illustration
Robotics illustration
Safety illustration
By the numbers

The human layer behind the AI you actually use

58K+
Vetted Experts
70
Countries
10K+
Licensed physicians
76
Languages
100%
On Payroll

Working with hospital systems, law firms, robotics labs, and the AI teams quietly shipping the next decade.

Notes From The Field

What We’ve Learned Building Human Data For The Teams Shaping AI.

Research

Pre-Labeling with Calibrated Confidence: Benchmarking Jev as a Rubric Judge and Dialect Pre-Labeler

Research

The Best Round Isn't Always the Last One

Research

Reading the Fine Print: Evaluating a Frontier Language Model on Saudi Legal Documents

AICPA SOC 2 Type II Certified

SOC 2 Type II Certified

Perle has achieved SOC 2 Type II certification, independently verified by a licensed CPA firm. Our security controls for availability, confidentiality, and data integrity are audited annually.