TL;DR

• AI project management can speed planning, risk management, and early design.
• Documentation and compliance demand human oversight to avoid errors.
• Clear communication strategies prevent wasted effort in “LLM-telephone” loops.
• AI supports, not replaces, creative thinking in innovation.
• Future PMs will stand out by combining AI fluency with judgment and clarity.

Project managers are on the front lines of rising complexity in medical device development. They sit at the intersection of vendor timelines, regulatory constraints, and engineering realities, all while juggling a growing suite of tools and environments chosen by other stakeholders.

It’s no surprise that many PMs are drawn to the promises of new AI-based systems and startups, which claim to help manage this growing problem and to allow users to cut through the noise to focus on the core skills of decision making, coordination, and communication.

Planning and Risk Management with AI

AI tools can be utilized to synthesize scattered information into clearer plans, and flag early-stage concerns. Consolidating varied information into executive summaries reduces decision latency, and enables more diverse sources to be used; however, offloading even part of the decision-making process to a black-box program can have serious downstream risks. If it is difficult to tie the decision to a specific piece of information, or if an important aspect of the data was ignored arbitrarily by the LLM, then trust in the process as a whole can start to erode.

The clearest mitigation to these risks is to only use LLMs when necessary, to minimize the opportunities for automated mistakes. The format of AI output may make intermediate results appear much tidier than they are in reality, so performing your own analysis and breakdown is key before presentation to stakeholders.

Documentation and Compliance in Medical Devices

Specialized tools, such as regulatory search engines purpose-built for Med Tech, can help simplify complex regulatory pathways and requirements. Identifying FDA product codes, relevant standards, and guidance documents can jump-start development with the right regulatory context and minimize wasted effort.

That said, regulatory documents demand extreme precision. AI-generated summaries and syntheses should be avoided, particularly in final products. How you utilize AI is important, and you must avoid the risk of hallucination – when generative models output information that is not based on fact, especially in this domain.

Communication and Collaboration Tools for PMs

AI and Large Language Models (LLMs) have created a small revolution in communication. They have clear utility in both summarization and in expanding on point-form notes, allowing for long messages to be summarized conveniently, and also for expansive brainstorming from sparse notes. When used thoughtfully, these tools can save a lot of time.

The fear, however, is that one stakeholder will use AI to expand their brief notes into a long email, which another stakeholder will summarize into brief notes. This game of LLM-telephone accomplishes nothing, when both would likely have been perfectly happy to exchange rough notes in the first place. What is needed are shared communication expectations, with the goal of eliminating complexity-compounding steps.

Accelerating Innovation and Early Design with AI

AI is routinely utilized to enable brainstorming sessions, expanding on rough ideas and prompting alternative approaches to problem solving. Used responsibly, LLMs amplify rather than replace creative thinking. With the right technical context, expertise, and judgment, they can accelerate progress toward refined concepts.

This requires the same willingness to step back as other fast-moving conceptual processes, however. An LLM cannot judge feasibility, and while the statistical nature of their construction can supercharge exploration of related concepts, a project manager must ensure that medical device development doesn’t get too bogged down in pie-in-the-sky ideation, at the cost of developing the real, feasible components of the system.

AI brainstorming may be best implemented after an initial proof-of-concept has been developed, and limitations of all aspects of the system, such as hardware, software, and AI/ML constraints have been discovered.

AI and the future skillset of Project Managers

As AI use becomes increasingly embedded into project workflows and the tools PMs use every day, PMs won’t be evaluated on whether or not AI is used, but how effectively it is harnessed to empower medical device development.

Sentiment analysis, communication triage, and real-time prioritization are key areas where AI can provide real value to project managers. For example, a system that can flag tone changes in client (or engineer) communications, or pre-emptively escalate issues based on collated project information could dramatically increase responsiveness in areas that clients care most about, and reduce downstream risk.

The most alluring aspect of AI is its claim to solve ever-growing complexity in all aspects of life. This is an illusion, however; complexity will always grow until the point its expansion is explicitly stopped.

Success will depend on recognizing that not all complexity is necessary, cutting to the core of an issue, instead of generating thousands of words that may never be read, or knowing when to sit down to read and understand a long document despite the allure of the “summarize this for me” button. These will differentiate those who will succeed versus those who will drown under the sea of new AI-powered tools.

As with other fields in the age of AI, the future AI-empowered project manager will:

• Use AI to inform, not decide
• Communicate with clarity and intent
• Bring deep insight, not just speed
• Look for opportunities to eliminate complexity, not just plaster it over

Related Resource