Integrations
DeepCog connects to your existing clinical infrastructure in hours, not months. Pre-built connectors for every major EHR, lab system, imaging platform, and data standard — with a REST API for everything else.
Works with the tools you already use
Integrate in minutes with our SDK
The DeepCog API is REST-native, OpenAPI-documented, and ships with Python and TypeScript SDKs. FHIR R4 resource bundles are accepted natively — no transformation layer needed.
- OpenAPI 3.1 specification with interactive Swagger docs
- Python, TypeScript, and Java SDK with async support
- FHIR R4 and HL7 v2.x natively parsed — no custom ETL
- Streaming responses for real-time clinical note generation
- Webhook support for async batch inference results
- Rate limiting and quota management per API key
# Initialize with your API key
client = DeepCogClient("dc_prod_xxxxxxxxxxxxxxxx")
# Send a FHIR R4 patient bundle
bundle = FHIRBundle.from_epic("patient_12345")
# Run clinical reasoning
result = client.infer(
model="ClinicalReasoner-13B",
input=bundle,
task="differential_diagnosis",
max_tokens=1024,
stream=True
)
# Stream the response
for chunk in result:
print(chunk.text, end="", flush=True)
# Result includes citations + confidence
print(result.citations) # [PubMed:38291...
Native support for every health data standard
FHIR R4
Full FHIR R4 resource support including Patient, Observation, Condition, Medication, DiagnosticReport, and DocumentReference. US Core and International Patient Summary compliant.
HL7 v2.x
Parse and route HL7 v2 ADT, ORU, ORM, and MDM messages. Automatic version detection from v2.1 through v2.8 with configurable field mapping.
DICOM
Ingest radiology and pathology studies directly from PACS systems. Our multimodal models process DICOM metadata alongside imaging reports and free text.
VCF / FASTQ
Upload Variant Call Format files for automated genomic variant annotation, ACMG classification, and clinical significance interpretation via GenomicLLM.
RxNorm & NDF-RT
Medication normalization and drug-drug interaction checking using the full RxNorm and NDF-RT ontologies, mapped to your formulary automatically.
SNOMED CT & ICD-10/11
Automated clinical coding with SNOMED CT, ICD-10-CM, ICD-11, and CPT — supporting diagnosis coding assistance and retrospective chart review.
Your stack, DeepCog-powered
Our solutions engineering team will map your integration requirements and have a proof-of-concept running within 48 hours.