On-premise ingest: manage a pynetdicom C-STORE service that assembles studies, applies backpressure, passes compressed transfer syntaxes, and spools to disk.
Cloud ingest: run a DICOMweb STOW-RS receiver with streaming multipart parsing, per-study UID checks, and one inference firing per request.
Tenancy and privacy: determine customer assignments, apply DICOM PS3.15 confidentiality rules, and maintain re-identification encryption keys the customer holds.

Outbound Lanes:

Support JSON callbacks, dictation-system autotext, and HL7 v2 ORU^R01 over MLLP shaped to customer interface engines.
Build per-study tracing, three-component health checks, and continuous integration for gateway repositories.

Security and Debugging:

Give precise answers on transport, retention, and incident handling during hospital IT reviews.
Debug structural issues like unbounded reads causing OOM-kills, or network rules returning mid-stream 403s, using packet captures and memory profiles.

Who We Seek:

Experience implementing protocols from specifications (e.g., DICOM, HL7) and running software in locked-down infrastructure.
Familiarity with PACS, VNA, teleradiology, or HL7 interface engines is helpful but not required.
Stack includes Python, FastAPI, pytest, pynetdicom, pydicom, AWS (S3, SQS, ECS/Fargate, etc.), Docker, and Terraform.

A2Z Radiology AI

A2Z builds clinical AI that reads alongside radiologists. Its abdomen-pelvis CT triage device is FDA-cleared and the first to simultaneously triage seven urgent conditions, backed by Khosla Ventures.

Apply for This Position