a2z Radiology AI

Solutions Engineer

• Own the clinical data path end to end, including how studies enter hospital systems and how results are delivered back
• Build and maintain on-premise DICOM ingest using a pynetdicom C-STORE service
• Assemble studies, handle senders that do not signal completion cleanly, apply backpressure, preserve compressed transfer syntaxes, and spool to disk during link failures
• Build cloud DICOMweb STOW-RS ingestion with streaming, bounded-memory multipart parsing
• Apply per-part study-UID checks and ensure one canonicalization and one inference firing per request
• Determine customer tenancy with no silent default
• Apply DICOM PS3.15 confidentiality rules and fail closed before data leaves the customer network
• Maintain an encrypted re-identification store under a customer-held key
• Implement outbound JSON callbacks, dictation-system autotext, and HL7 v2 ORU^R01 over MLLP
• Adapt integrations to each customer’s interface engine
• Provide per-study tracing and three-component integration health monitoring
• Check drift between expected and deployed configurations
• Bring both gateway repositories under continuous integration during the first week
• Participate in hospital IT and clinical security reviews
• Provide precise documentation and answers about transport, retention, incident handling, and system behavior
• Debug integration failures using packet captures, memory profiles, and protocol specifications
• Remain reachable during the customer’s clinical day and travel occasionally for on-premise installations

Integration Engineer, Clinical Data Path

Platform Engineer

• Own the clinical imaging viewer built on Cornerstone3D and VTK.js
• Build a unified volume-rendering path with stack-rendering fallback
• Generate on-demand multiplanar reformats in any plane and label them clearly
• Implement progressive loading from the first approximately 10% of each HTJ2K codestream
• Ensure radiological left/right correctness, rulers under gantry tilt, MONOCHROME1 inversion, and signed-pixel codec compatibility
• Build and improve the 3D annotation system, including mask storage, click-to-segment across three planes, region growing, FWHM thresholding, and multi-seed refinement
• Develop annotation evaluation infrastructure including case-pool ledgers, annotator phase state machines, ground-truth and peer-overlap batches, agreement scorecards, and cross-series resampling through NIfTI affines
• Design per-lesion-category annotation schemas with gating by view, phase, and slice, including conditionally required fields, in collaboration with radiologists
• Build versioned sparse-RLE mask contracts keyed by SOP Instance UID
• Implement geometry mapping from model grids through image position, direction cosines, and pixel spacing into world-space contours
• Develop the FastAPI platform on AWS
• Enforce per-study authorization on every read and isolate customer cohorts
• Implement access-trail auditing meeting HIPAA §164.312(b) for study-data reads
• Improve annotation interfaces by observing radiologist workflows and increasing annotation throughput and model quality
• Collaborate with radiologists on clinical annotation decisions

Integration Engineer, Clinical Data Path

Clinical Imaging Platform Engineer