AI Solutions Engineer

At Holy Cross, we are focused on creating a world-class workforce by attracting and retaining employees dedicated to fostering a culture of professionalism, pride, mutual trust, and respect—one that values appreciation and recognition, empowering each of us to flourish both individually and collectively. We seek candidates who are inspired by our mission and values, and who are eager to make a meaningful difference on our campus. Learn more about Life at Holy Cross here. The AI Solutions Engineer builds and supports AI-powered applications, agents, and integrations that serve teaching, learning, and College operations.

The emphasis of this role is agentic: designing and building systems that plan, call tools, and carry multi-step work through to completion with appropriate human oversight, rather than systems that simply answer questions. The Engineer works closely with the Principal AI Solutions Engineer, who sets architectural direction and owns technical strategy. Within that direction, this role turns defined problems into working, instrumented, supportable software, releasing in small increments. It is an individual contributor role with no supervisory responsibility, expected to grow toward greater architectural ownership over time.

The Engineer also partners with the Director for Applied Artificial Intelligence and Emerging Technologies, so that ideas from faculty work have a clear path into working systems.

Job Description

Major Areas of Responsibility Agentic development Build, test, and deploy AI agents that plan, select and call tools, and complete multi-step tasks, with human-in-the-loop checkpoints where judgment or risk requires them. Implement the tool and function interfaces that let agents act on College systems safely, using modern APIs and middleware. Design agent workflows that degrade gracefully: bounded scope, clear permissions, sensible retries, and a defined fallback when a task cannot be completed. Develop point solutions, applications, and integrations using agent frameworks such as the Claude Agent SDK and Google ADK.

Translate requirements into working software quickly, favoring small releases over long build cycles. Evaluation, guardrails, and reliability Build evaluation sets and test harnesses so agent behavior is measured rather than assumed, and regressions surface before users do. Implement guardrails, permission scoping, and logging so that agent actions are auditable and failures are contained. Monitor deployed agents for cost, latency, error rates, and behavioral drift, and tune accordingly. Support and maintain existing applications and agents in production.

Collaboration and delivery Work closely with the Principal AI Solutions Engineer, following architectural direction and contributing to design and technical review. Partner with the Director for Applied Artificial Intelligence and Emerging Technologies on faculty-facing work and on the AI Exploration Lab. Coordinate with colleagues in Enterprise Systems…

Read the full posting