Skip to content
syda
Paper records for a patient, visit, diagnosis, and treatment connected by one green thread, with each record referencing the one before it.

Synthetic data.Real-world logic.

Describe the world your data lives in. Syda generates connected records that follow its rules, without using customer data.

Follow the connections.

A patient has visits. A visit has a diagnosis. A diagnosis has a treatment. The meaning lives in how they connect.

Select a record to inspect its fields and follow its links. Switch patients to explore a different care path.

Patient careInteractive synthetic scenario
Patient
Visits
Diagnosis
Treatment
Routine follow-up5 records4 linksAll references resolve

From brief
to dataset.

Real systems have relationships, decisions, and exceptions. Your test data should, too.

Describe the system.

Start with the entities, lifecycle, and edge cases. Syda turns your brief into an editable scenario.

Your scenario

“Create patient care data with visits, diagnoses, and treatments. Include routine follow-ups and escalated care.”

PatientVisitDiagnosisTreatment
Four entities, linked by their IDs.

Keep the logic.

Refine schemas, paths, and constraints before generating. Define how records behave together, beyond individual fields.

Scenario constraintsPatient care
  • Every visit belongs to a patientRelationship
  • Follow-ups happen after the first visitSequence
  • Escalated care includes a specialist visitCare path

Inspect. Then export.

Preview your tables and review quality checks. Export CSV or JSON when the data is ready for your tests.

Sample relationship checks4 / 4 passed
  • Patient referencesPassed
  • Visit referencesPassed
  • Diagnosis referencesPassed
  • Unique record IDsPassed

Before
you build.

Does this use customer records?

You can start with a description of your system and its business rules. The generated records are synthetic, so you do not need to upload real customer records to build a scenario.

Can I edit the scenario?

Yes. Review and refine the tables, fields, relationships, workflow paths, and constraints before you generate the dataset.

What makes the data connected?

Records are generated with relationships across tables. A visit references a patient, a diagnosis references a visit, and a treatment references a diagnosis, keeping the scenario's context intact.

How do I check the output?

Preview the generated tables, inspect quality checks, and review the evaluation report before using your data. Your rules stay visible throughout the workflow.

What can I export?

Download generated tables as CSV or JSON. You can also export the evaluation report as JSON or HTML for review.

Is this demo live generation?

This page uses a small, curated synthetic patient-care sample to demonstrate linked records. Create a workspace to design your own scenario and run generation.