Skip to Content

Pre-modeled
ontologies.

Generate test data from industry ontologies that already know every object, link and rule.

Generate a scenarioRead the docs
Pharma and medtech quality
28 objects, 41 links
Batch B2024001business code batchNumber
Deviation batchId1:N
Complaint batchId1:N
Investigation INV-2024-001business code investigationCode
Capa investigationId1:N
ActionItem capaId1:N
Zero orphaned rows. The highlighted chain shares one batch and closes on one CAPA.

Two industries are modelled. Yours can be the third.

Pharma and medtech

Batch genealogy, deviations, complaints, investigations, CAPA and change control, tracked through to closure.

28 objects, 41 links
@shpitdev/prismantix-ontology-operations

Healthcare payer and provider

Payers, plans, members, coverage and providers, plus the full eligibility inquiry and response chain.

8 objects, 13 links
@shpitdev/prismantix-ontology-healthcare

Or bring your own

Point the CLI at a Foundry ontology export. It comes back as a package you can build on.

prismantix plugins install @shpitdev/prismantix-ontology-interop
prismantix foundry import \
  --input ./ontology-export.json \
  --output ./foundry-oac \
  --validate

The schema travels with the rows.

Palantir Foundry

Available

Convex

Available

Turso DB

Available

CSV bundle

Available

Databricks

Planned

Snowflake

Planned

PostgreSQL

Planned

A CLI and an endpoint.

No SDK to adopt.

Command line

Queue an export, poll it, get the files.

npm install --global @nyrra/prismantix-cli
export PRISMANTIX_API_KEY="sk_..."

prismantix records pull \
  --client LLY \
  --output ./prismantix-exports \
  --json

Chat completions endpoint

OpenAI-compatible. Point any harness at it and ask for data in words.

POST https://llm-proxy.prismantix.com/v1/chat/completions
Authorization: Bearer $PRISMANTIX_API_KEY

{ "model": "gpt-5.6-luna",
  "messages": [{ "role": "user",
    "content": "Profile the pharma quality ontology" }] }

Works with OpenCode today, and anything else that speaks chat completions.

Start with the pharma ontology.
Generate a scenarioRead the docs