Senior AI and Decision Engineer

What matters most

• Build decision support around operational needs

  The role calls for analyzing workflows and operational needs, then developing AI-enhanced decision systems and software prototypes, including LLM capabilities. For preparation, focus on how a decision task becomes a useful prototype rather than on a model alone.

• Evaluate solutions for consequential decisions

  MITRE specifies designing, evaluating, and fielding AI solutions, assessing them against sponsor requirements, and producing actionable recommendations. Prepare to explain how evaluation results would support a recommendation or reveal a gap.

• Advance research while leading delivery

  The role includes proposing and executing research in AI-enhanced decision systems, identifying new R&D opportunities, and leading projects or small teams. Prepare examples that distinguish a research contribution from its practical decision-support value.

• Work across disciplines and audiences

  Cross-functional teamwork is a stated responsibility, and communication with both lay and expert audiences is preferred. Prepare to explain complex technical choices in terms of the decision problem and operational needs.

Likely challenges to think through

• Balancing model capability with decision reliability

  A likely difficulty is deciding when an LLM or other AI capability is useful enough for a consequential decision task. Prepare to discuss limitations, testing, and how you would identify gaps before recommending use; this is a preparation hypothesis, not a reported MITRE problem.

• Turning a prototype into a fieldable solution

  The posting spans workflow analysis, prototyping, evaluation, and fielding. A likely trade-off is moving quickly on a prototype while still meeting sponsor requirements. Prepare to explain what evidence would justify further development versus a narrower recommendation.

• Keeping research relevant across problem domains

  Novel research and work on varied sponsor decision problems may pull in different directions. Prepare to explain how you would select a research question with both technical merit and a clear connection to an operational need; the posting does not prescribe a selection process.

Topics to prepare

• Machine learning, deep learning and reinforcement learning

  Required hands-on areas. For revision, compare when a supervised model, deep model or reinforcement-learning approach would fit a decision problem; identify the data, feedback signal and failure modes for each. This comparison is a preparation exercise, not a stated interview requirement.

• AI-enhanced decision systems

  The role develops systems for consequential sponsor decisions. For revision, take a hypothetical operational workflow and map its decision points, users, constraints and where AI output could inform rather than replace judgment. This is a suggested exercise.

• Large language models in decision support

  The posting calls for LLM-based capabilities and prefers recent LLM experience. For revision, outline a plausible LLM-assisted decision-support use case, including how you would check its outputs against the decision need. The use case is suggested practice, not an assigned project.

• AI evaluation and high-stakes limitations

  The role evaluates solutions against sponsor requirements and identifies gaps in high-stakes AI decision making; AI testing and evaluation is also a listed technical area. For revision, draft criteria for assessing usefulness, errors and limitations in a hypothetical decision system. The criteria are a preparation suggestion.

• Research opportunities in decision systems

  The role proposes and executes novel research and identifies new R&D opportunities. For revision, frame one hypothetical gap as a research question, then distinguish what an experiment could establish from what would remain uncertain. This is suggested practice.

Skills to revise

• Programming in Python or C++

  Experience with languages such as Python and C++ is listed. For hands-on revision, implement a small decision-support component in a language you know and practice explaining its inputs, outputs and tests. This exercise is suggested, not specified by MITRE.

• Workflow-to-prototype development

  The role builds decision-support software prototypes from workflow and operational analysis; prototype development and deployment are preferred experience. For revision, sketch a minimal prototype from a hypothetical workflow, then implement one working component and state what would still be needed to field it. This is suggested practice.

• Requirements-based evaluation and recommendations

  The role evaluates AI solutions against sponsor requirements and delivers actionable recommendations. For revision, write a short evaluation of a hypothetical prototype against explicit requirements, including evidence, limitations and a recommended next step. This is a suggested exercise.

• Cross-functional technical communication

  The role works across disciplines and leads projects or small teams; communication for lay and expert audiences is preferred. For revision, explain the same hypothetical decision-system trade-off in a brief technical summary and a plain-language summary. This is suggested practice.

• Build a requirements evidence sheet

  The posting specifies degree-and-experience paths, hands-on machine learning including deep learning and reinforcement learning, at least one additional listed specialty, and experience with languages such as C++ and Python. Prepare a one-page index of examples and artifacts you could discuss for each.

• Rehearse a decision-system project walkthrough

  The role involves analyzing workflows and operational needs to build decision-support prototypes. If you have a relevant project, prepare a concise walkthrough of the decision, users, prototype, your contribution, and what changed after testing; otherwise, use a clearly labeled hypothetical design.

• Prepare a technical design exercise

  Sketch how you would turn a high-stakes decision workflow into an AI-enhanced prototype, including where an LLM might help, where human judgment remains necessary, and how you would identify gaps. Present these as design choices to discuss, not as MITRE requirements.

• Draft an evaluation and recommendation brief

  The posting calls for evaluating AI solutions against sponsor requirements and producing actionable recommendations. Prepare a sample plan with success measures, failure cases, testing approach, and a short recommendation tied to the results.

• Collect research and leadership examples

  Prepare one example each of proposing an AI research direction and leading a project or small team, if available. Note the problem, collaborators, decisions you made, outcome, and how you communicated technical tradeoffs to different audiences.

• Check practical requirements before discussing logistics

  Review the stated onsite expectation of at least three days a week and the listed McLean and Bedford locations. Be ready to discuss the clearance timeline accurately: none is required at application, while the hired candidate must have or obtain Secret clearance within one year; U.S. citizenship is required for clearance eligibility.

• Write focused questions for the employer

  Prepare questions about the decision workflows this team supports, how sponsor requirements shape prototype evaluation, what fielding an AI solution entails, and how research opportunities and small-team leadership are prioritized.