Skip to main content
A violet passionflower among dark leaves
Pasiflora AI
For teams training domain-specific AI

Expert data for the AI you're building

PhDs, MDs, and senior practitioners producing the training, evaluation, and red-team data your domain model gets judged on. A pre-vetted network means a project can start in days, not after months of hiring.

Tell us about your projectSee how matching works
4,410+specialist fields covered
135countries represented
8roles hiring right now
$35–$250per hour on assigned work, by domain
Standards we hold to
  • CredentialedPhD, MD, or senior practitioner
  • Bloom-trainedbefore any billable task
  • Human-reviewedevery submission, by more than one reviewer, before delivery

Rates vary by domain, task complexity, and client demand. Approval to the expert pool does not guarantee task volume.

Latest roles

8 open now · 133 fields listed

View all roles
i. What we do

Three ways to put credentialed experts on your model

Expert data projects

PhDs, MDs, and senior practitioners producing the training, evaluation, and red-team data your domain model gets judged on.

  • Annotation, evaluation, generation, validation, comparison
  • Scientific, medical, and financial AI
  • Scoped paid pilot before you commit to scale
Tell us about your project

Tiger Teams

Five to ten credentialed experts deployed in days, not weeks. Built to rescue backlogs, clean quality drift, or ship high-volume work on accelerated timelines.

  • Backlog rescue
  • Quality cleanup
  • High-volume sprint
Talk to us about a deployment

AI Agent Testing

Continuous testing for customer-facing AI agents. Simulated customers probe your chatbot on a schedule; credentialed experts review the findings that matter.

  • Scoping call
  • Continuous testing
  • Expert review and reporting
Book a scoping call
ii. Verticals

Built for the verticals general crowdsourcing can't serve

Domain models live or die on the quality of expert data behind them. We staff your project with practitioners whose credentials map to the questions your model has to answer.

Scientific AI

PhD researchers and senior engineers producing graduate-level reasoning data, eval sets, and adversarial prompts for models working in science, math, and software.

Eval setsPreference dataRed-team promptsCitation grounding

Medical AI

Board-certified physicians and clinicians grounding diagnostic, clinical-summary, and patient-facing models in evidence and standard of care.

Clinical evalRAG groundingDifferential reasoningSafety review

Financial AI

CFAs, accountants, and senior analysts producing the numerical-reasoning, filings-analysis, and compliance data that finance models actually get tested on.

Filings analysisNumerical reasoningComplianceMemo generation
Also active in scientific research, engineering, code review, and 4,405+ other specialist fields.
Fields listed now
  • Remote English Customer Caller
  • Customer Service Agent – U.S. Accent Required
  • Multi-App Data Project
  • Explosives AI Safety Evaluators & SMEs
  • Image-to-Code / Visual Reconstruction Engineer
  • Radiological AI Safety Evaluators & SMEs
  • Nuclear AI Safety Evaluators & SMEs
  • Chemical AI Safety Experts
  • Technical Decision Documents
  • Sales & Business Development Decks
  • Public Policy Analyst
  • Urban Planning Expert
  • Emergency Management Specialist
  • Real Estate Appraiser
  • Defense & Intelligence Analyst
  • Film & Video Production Expert
  • Music Theory & Composition Expert
  • Game Design Consultant
  • Graphic Designer & Brand Expert
  • Photography & Visual Arts Expert
  • Journalism & Editorial Expert
  • Agronomy & Crop Science Expert
  • Food Scientist
  • Culinary Expert & Chef
  • Insurance Underwriter
  • Property Manager
  • DevOps & SRE Expert
  • Mobile App Developer
  • Database & SQL Expert
  • Prompt Engineer
  • QA & Test Automation Expert
  • Dermatology Expert
  • Dental Expert
  • Emergency Medicine Physician
  • Anesthesiology Expert
  • Speech-Language Pathologist
  • Marine Biologist
  • Geologist & Earth Science Expert
  • Statistician & Biostatistician
  • Investment Banking Analyst
  • Auditor & Forensic Accounting Expert
  • Aviation & A&P Mechanic
  • Carpenter & Cabinetmaker
  • Family Law Expert
  • Tax Law & Corporate Counsel
  • Translation
  • Indigenous Domain Expert
  • Pharmacy Technician
  • Medical Coder / Biller
  • Mental Health Counselor
  • Physical Therapist
  • Registered Nurse
  • Tax Preparer
  • Personal Finance Coach
  • Bookkeeper / Accounting Specialist
  • Regulatory Compliance Analyst
  • Paralegal Specialist
  • ESL / Language Specialist
  • Corporate Trainer
  • K-12 Teacher / Curriculum Reviewer
iii. Benchmarks

We publish how the frontier actually performs

Pasiflora builds evaluations that separate real capability from recall, and reports them honestly, artifacts and limits included. Each benchmark runs live across the frontier labs.

All benchmarks

Live SQL over a messy production database · 12 questions · ground-truth graded

  1. 01Claude Opus 4.8 · Anthropic91.7%
  2. 02Claude Sonnet 4.6 · Anthropic83.3%
  3. 03GPT-5.2 · OpenAI72.2%
  4. 04Gemini 2.5 Pro · Google66.7%
  5. 05Claude Haiku 4.5 · Anthropic61.1%
questions correct · updated July 12, 2026Full leaderboard and method
iv. How a project runs

From first conversation to delivered data

Every project starts with a direct scoping call, because cookie cutter data pipelines don't work for specialized domains.

01

Scope

Tell us the domain, volume, format, and quality bar. We follow up within one business day to scope it together.

v. Before a pilot

Questions every buyer asks before a pilot

Straight answers on ownership, security, and how a first project works.

Your call, and it's set in the engagement. Most clients choose exclusive ownership: deliverables and derived IP transfer to you on delivery and we don't reuse or license them to anyone else. If exclusivity isn't a priority, we can scope a shared-rights arrangement instead.

Most engagements begin with a scoped paid pilot so you can validate quality before committing to scale.

Every expert placed on a project is credentialed and Bloom-trained before any billable task, and every submission passes through more than one reviewer before delivery.

vi. From the blog

Latest writing

All posts
Enterprise · May 14, 2026

When Your Data Campaign Is on Fire, You Don't Need a Bigger Crowd. You Need a Tiger Team.

When a frontier AI campaign starts to fail, throwing more contractors at it makes things worse. Here's how Pasiflora's Tiger Teams work and when to deploy one.

7 min read
Company · Apr 2, 2026

Building the Human Layer for AI

Why high-quality human data is the foundation of frontier AI systems, and how we're building the infrastructure to deliver it.

6 min read
AI Research · Apr 7, 2026

How Domain Experts Shape AI Alignment

The critical role that PhDs, doctors, and specialists play in making AI systems safer and more accurate.

5 min read
A dusk sky over distant hills

Tell us about your project

Share the domain, the task type, the volume, and the timeline you have in mind. We'll review it and get back to you within one business day to scope a credentialed expert pool and a path to a pilot.

Tell us about your projectOr talk to a human · clients@pasifloraai.com
For experts

Have a PhD, MD, JD, or deep domain expertise?

Apply to the network. Premium rates, real credentials, and work that shapes how frontier models behave.

Apply as an expertSee open roles