Design narrow adapters for Adobe SDKs and REST APIs with pinned versions, typed contracts, secret and URL redaction, and bounded polling for asynchronous jobs. It also helps record compatibility evidence and migration notes before approved live actions.
Map and review an Abridge clinical documentation workflow from consent and recording through draft review, evidence checks, edits, and final disposition. Keep clinician approval explicit, identify failure exits, and separate verified facts from tenant-specific details and pending approvals.
Prepare a bounded support bundle for an Abridge issue using allowlisted operational details such as versions, coarse timestamps, and correlation references. Exclude recordings, transcripts, patient identifiers, credentials, and unrestricted logs, then document redactions and review details.
Review access across Adobe organizations, projects, product profiles, and technical accounts. The skill maps effective permissions, identifies stale or overly broad access, and prepares an approved, reviewable plan for provisioning, revocation, and periodic recertification.
Audit how clinician-reviewed Abridge note sections map into Epic templates and downstream workflows. Use approved test records to document destinations, exceptions, evidence, rollback steps, and a go or no-go decision for a specific environment.
Review Adobe integrations for approved endpoints, versions, data handling, spending limits, and approval requirements. Produce clear allow and deny rules, enforcement points, tests, and evidence so teams can check risky actions before they run.
Investigate Adobe API failures using sanitized responses, configuration, and service documentation. Distinguish authentication, permissions, request validation, throttling, and service issues, then document a limited corrective action, retry decision, verification, and evidence for unresolved support cases.
Review data handling across Adobe workflows, including classification, minimization, storage, signed URL custody, retention, deletion, content provenance, and privacy requests. The skill produces a reviewable record of data flows, decisions, tests, owners, and unresolved gaps.
Prepare a staged Abridge pilot or expansion by organizing cohort scope, care setting, training, support, success measures, and rollback criteria. The skill produces a readiness checklist and decision handoff; it does not deploy Abridge infrastructure.
Add a synthetic-data CI gate for customer-owned Abridge integration code. Test consent handling, note-review handoffs, redaction, and rollback behavior with mocked vendor boundaries, without connecting CI to a live Abridge or EHR tenant.
Design service-specific controls for Adobe workloads, including request admission, Retry-After handling, capped retries, and queues with spend limits. It helps respond to 429 errors and scaling pressure using verified service constraints and test evidence.
Adapt pretrained models to new datasets and tasks through transfer learning, which reuses existing training. Generate Python code for fine-tuning, data preparation, validation, and evaluation, then save the adapted model, training logs, metrics, and documentation.
Plan a controlled Abridge workflow migration by documenting source and target states, affected settings and cohorts, test evidence, cutover steps, and rollback conditions. It helps teams base changes on tenant-specific release evidence and approved test procedures.
Map how Abridge connects with electronic health records, identity systems, devices, and clinical review. Produce architecture diagrams and data flows that identify responsibility boundaries, failure paths, verified evidence, and unresolved interface assumptions.
Prepare an Abridge implementation by mapping tenant access, identity owners, roles, and EHR integration requirements to approved evidence. The result is a readiness inventory with verified facts, unresolved gaps, and actions that still need approval.
Deploy Adobe-backed services or App Builder apps with separate environments, verified build artifacts, and limited trial traffic. The workflow checks monitoring and rollback readiness, then records evidence for release decisions.
Measure latency and throughput across Adobe integration steps, identify bottlenecks, and compare bounded improvements. The workflow tracks output correctness and spend alongside speed, then records evidence, the selected change, and rollback thresholds.
Choose an Adobe authentication flow based on who owns the data, confirm the required entitlement and product profile, and document secret handling. The workflow yields a reviewable setup plan and a read-only proof before integration work proceeds.
Coordinate Adobe Firefly image generation from prompt and budget approval through job monitoring and artifact storage. Track returned status and cancellation links, validate generated files, preserve provenance metadata, and document failures, timeouts, and approved cancellations.
Review an Adobe integration before launch or a major change. This skill checks authentication, entitlements, API versions, data handling, asynchronous reliability, spending, monitoring, and rollback, then records a go, conditional go, or no-go decision with evidence.
Verify a single Adobe API connection before building a production workflow. Choose a current, low-impact read or synthetic operation, check the required entitlement, and create a redacted receipt that records the outcome without exposing secrets or customer content.
Design recovery controls for asynchronous Adobe workloads, including duplicate prevention, time-limited status checks, service-specific queues, and failure isolation. Produce state models and recovery runbooks that help teams investigate uncertain jobs before approving a replay.
Run an approved Adobe PDF Services job for document conversion, extraction, or generation. Track asset custody and asynchronous status, verify the returned file, and record cleanup and retention evidence for the operation.
Plan capacity, queues, and retry rules for an Abridge integration using tenant-specific contracts and verified responses. Define how to handle overload, prevent duplicate clinical actions, and hand off work when processing cannot continue.
Is this your account? Link GitHub to bring it onto your creator page (Japanese)