Integration Engineer, Clinical Data Path

Role overview

An integration engineer is needed to own the end-to-end clinical data path that connects hospital imaging systems to an AI triage model and delivers results back into radiologist workflows. The work centers on the DICOM and HL7 plumbing that determines whether a cleared medical device actually functions inside a real hospital network. This is a customer-facing engineering role that pairs deep standards knowledge with practical on-premise deployment experience.

Responsibilities

- Operate and harden an on-premise `pynetdicom` C-STORE service that assembles studies, tolerates senders that never signal completion cleanly, applies backpressure, preserves compressed transfer syntaxes, and spools to disk so nothing is lost when network links drop. - Maintain a cloud DICOMweb STOW-RS receiver built around a streaming, bounded-memory multipart parser with per-part study-UID checks and exactly one canonicalization and inference firing per request. - Resolve customer tenancy at the network boundary with no silent defaults, apply DICOM PS3.15 confidentiality rules, run fail-closed checks before data leaves the customer environment, and keep the re-identification store encrypted under a customer-held key. - Shape outbound result delivery to whatever interface engine each hospital runs, including JSON callbacks, dictation-system autotext macros, and HL7 v2 ORU^R01 over MLLP. - Provide per-study tracing, three-component integration health checks, drift detection between intended and actual deployments, and continuous integration coverage for gateway repositories. - Participate directly in hospital IT and clinical security reviews, respond during customer clinical hours, and travel occasionally for on-premise installs in a locked-down hospital environment.

Requirements

- Hands-on experience with DICOM protocols (C-STORE, DICOMweb STOW-RS) and HL7 v2 messaging, including MLLP transports. - Familiarity with DICOM PS3.15 confidentiality profile and de-identification patterns relevant to medical imaging data. - Comfort building streaming, bounded-memory parsers and engineering for failure modes such as partial transfers, dropped associations, and ambiguous sender behavior. - Ability to debug cross-stack integration issues where network, TLS, firewall, and application layers produce misleading symptoms. - Willingness to join customer-facing security and IT reviews and to be reachable during the customer's clinical day. - Location must match the account base: US-based for US on-premise customers, or regional for international deployments, and the work location is a disclosed sub-processor location in customer agreements.

Benefits and work setup

- US base compensation of $140,000 to $180,000 plus an approximately 10% discretionary bonus, with equity possible for top candidates. - International hires are remote within their home country, paid in cash-weighted local-market bands shared early in the process, with no visa or immigration sponsorship required and no equity component. - Customer-driven placement rather than a free choice of remote location, given the customer-facing and sub-processor constraints described above. - Occasional travel required for on-premise installations.