DEV Community

Manohar Papasani
Manohar Papasani

Posted on

How do you align with stakeholders while protecting your team?

Asked in an AI Engineering Manager interview.

What they are really testing: whether you can run stakeholder demand without letting every request land on your engineers, and whether you can say no in a way that keeps trust intact.

A strong answer:

  • Set up a single intake channel for AI work, with a written template that asks for the business outcome, the user, the success metric and the deadline. Requests that skip it wait in a queue, which makes priorities visible.
  • Publish a roadmap that is reviewed on a fixed cadence with product, platform, legal and research leads, so stakeholders see what is committed, what is next and what is parked.
  • Make trade-offs explicit. When capacity is full, show the cost of each new ask in terms of what it displaces, and let the stakeholder choose the order rather than absorbing the cost silently.
  • Say no with reasons tied to criteria already agreed (risk, expected value, effort, dependencies), and offer a path back in, such as a later slot or a smaller scoped spike.
  • Protect the team's time explicitly: define on-call and support rotations, limit how many concurrent workstreams one engineer carries, and treat interrupts from executives as roadmap changes that go through the same intake.
  • Revisit decisions when the facts change. If a new model or a managed service makes custom work unnecessary, say so early and retire the work, rather than defending the original plan.

Common mistakes:

  • Saying yes to everything to look collaborative, then pushing the overload onto the team without a visible trade-off.
  • Treating stakeholder alignment as a series of one-to-one meetings with no shared priority list, so each group leaves with a different story.
  • Protecting the team by hiding work from stakeholders, which breeds distrust when delivery slips and nobody saw the constraints coming.

How would you structure the intake and prioritisation process for a team serving five or six stakeholder groups at once? Share your approach below.


Originally shared on LinkedIn.

An AI interview question every day, leadership for tech people and hands-on AI learning:

Top comments (0)