Artificial intelligence can change a medical device after it reaches the market. For employers, that makes one question essential before opening a search: who owns each decision when the software, the device, or its intended use changes?

Image: ThisisEngineering/Unsplash.
The FDA’s final guidance on Predetermined Change Control Plans (PCCPs) for AI-enabled device software describes information a marketing submission should provide about planned modifications, the methods for developing, validating and implementing them, and an assessment of their potential effects. The guidance does not prescribe an organization chart or require a particular job title. It does give employers a concrete way to examine whether responsibilities are clear across product, engineering, quality and regulatory teams.
Start with the change pathway
Before defining a vacancy, map how a proposed software or model change moves from an idea to a controlled release. Identify who proposes a change, who evaluates its effect on the device and users, who designs and tests it, who reviews the evidence, and who authorizes implementation. Also identify who monitors performance after release and who can pause or escalate a change when results differ from expectations.
The answers may involve several functions. Depending on the product and the company’s structure, responsibilities could sit with product management, software or machine-learning engineering, systems engineering, quality, regulatory affairs, clinical or human-factors specialists, data teams, or cybersecurity. These are possibilities to assess—not a universal staffing checklist. For related role-scoping examples, see our guides to computer software assurance in medical devices and medical-device cybersecurity hiring.
Turn responsibilities into a hiring brief
For each handoff, specify the work the person will own and the decisions they can make. “AI experience” by itself is too broad to define a role. A stronger brief describes the product context, the change process, the evidence expected, the partners involved, and the boundary between the role’s authority and the organization’s approval process.
For a software or machine-learning position, for example, clarify whether the person will develop or evaluate models, maintain data and testing workflows, document version changes, or support deployment and monitoring. For quality or regulatory positions, define whether the work includes reviewing change impact, coordinating records, assessing submission implications, or advising teams on established procedures. Do not assume that one hire should own all of these functions.
Questions to settle before recruiting
Hiring teams can make a search more precise by answering a few practical questions:
What is changing? Describe the software function, product, user setting and types of modifications the team expects to consider.
What evidence will the role produce or review? Define the testing, validation, documentation, risk assessment or monitoring work that belongs to the position.
Where are the handoffs? Name the partners who develop, evaluate, approve, release and monitor a change. Identify who resolves disagreements and escalates unresolved risk.
What must the person have done before? Separate essential experience from trainable knowledge. Specify relevant product, software lifecycle, regulated-device, quality or cross-functional experience only when it is tied to actual duties.
Assess candidates against the work
Use a consistent, job-related assessment. Ask candidates to explain how they would clarify ownership for a proposed change, identify the evidence and stakeholders needed for review, and communicate a concern that could affect release timing. A structured scenario can reveal how a candidate reasons across technical and quality boundaries without asking them to disclose confidential information from a former employer.
Agree on evaluation criteria before interviews begin. For example, the panel might assess whether the candidate identifies affected functions, distinguishes technical testing from approval authority, recognizes the need for documented decisions, and communicates uncertainty clearly. Apply the same criteria to each candidate and connect every score to a defined responsibility in the role.
Build the team around accountable handoffs
A PCCP is a regulatory submission element, not a substitute for an employer’s internal governance or a guarantee that a proposed change will be authorized. The FDA’s guidance is useful here because it prompts teams to define planned modifications, methods and impact assessment. Employers can use those topics to clarify work and hiring needs while seeking regulatory advice for product-specific decisions.
When the responsibilities are clear, a search can focus on the actual gap: technical development, validation, quality oversight, regulatory coordination, or leadership across those functions. HealthStaffingGroup supports permanent placement, contingency search, retained and executive search, and Recruiting Subscription Service for employers hiring medical-device and MedTech talent.
Sources
FDA, Marketing Submission Recommendations for a Predetermined Change Control Plan for Artificial Intelligence-Enabled Device Software Functions (final guidance, August 2025).