Flock by Pidgeon Health

Test databases that look like a real hospital

10,000 patients with realistic diagnoses, encounters, and relationships — but none of them are real. Flock generates synthetic populations that catch the same edge cases production throws, without a single line of PHI.

Join the Early Access Waitlist Download Free CLI

The Difference Between Testing and Guessing

Sanitized production copies miss edge cases. Random data misses everything. Flock generates populations grounded in real-world demographics — Census, CDC, and clinical correlation data — so your tests find what production will throw.

Schema Intelligence

Point Flock at your database and it figures out the rest — classifies healthcare tables, maps relationships, and learns your schema structure automatically.

Realistic Populations

Patients with real-world demographics, correlated comorbidities, temporal coherence across encounters, and family linkage. Not random noise — populations that test like the real thing.

Data Learning

Ingest a sample from your existing data and Flock learns the patterns — distribution shapes, value frequencies, field correlations. Your synthetic data matches your data profile.

Every Format You Need

Output as FK-ordered SQL INSERT, CSV, HL7 message streams, or FHIR Bundles. One population, every format your team needs — seed a database or feed an interface.

Relational Consistency

Foreign keys resolve. Temporal ordering holds. Every row connects correctly across tables, encounters, and time. No broken references, no orphaned records.

Population Analytics

Verify before you seed. Demographic breakdowns, comorbidity distributions, and coverage statistics let you confirm your synthetic population looks right.

One Command, 10,000 Patients

Seed an entire test environment in minutes. Referential integrity verified.

Terminal

$ pidgeon flock generate \
    --schema postgres://localhost/ehr_test \
    --population 10000 \
    --output sql

Analyzing schema... 47 tables classified
Generating 10,000 patients...
  Demographics: Census 2020 ACS distribution
  Comorbidities: NHANES correlation matrix
  Encounters: temporal coherence validated
  Family links: 2,847 family groups

Output: ./flock-output/seed.sql
  INSERT statements: 142,391
  Tables seeded: 23
  Referential integrity: verified

Get Early Access

The first 100 members lock in founding pricing and shape what we build next. Your healthcare data deserves better tools — get them before launch day.

Founding Member Perks

✓30-day Console trial (vs. standard 14-day)
✓$29/mo locked for life (vs. $39/mo)
✓Priority Loft beta access
✓Direct access to the product team
✓"Founding 100" badge on your account