Build With AI, Not Just About It: Wentworth Institute of Technology’s Approach to Hands-On Faculty Development
Artificial intelligence now sits on nearly every higher education agenda in the USA, yet most faculty development efforts stop short of what instructors actually need. Institutions host awareness seminars and circulate policy memos, but few give faculty the structured time, expert guidance, and peer support required to build something concrete with AI. Wentworth Institute of Technology, a technology-focused university in Boston, Massachusetts, took a different path, and the results of its first cohort offer a practical model for campuses across the country.
When Josh Larson stepped in as director of digital transformation and academic technology services at Wentworth Institute of Technology last summer, he devoted his first months to listening. He met with deans and faculty across all five of the university’s schools, and in nearly every conversation the same request surfaced. “What I heard the most was, ‘I know about AI, but how do I actually do stuff with it?'” Larson said.
That question became the Faculty AI Academy, a cohort-based program run through ATLAS — Advancing Teaching, Learning, and Scholarship — the university’s teaching and learning hub. Eleven faculty members joined the 14-week pilot, spending the first seven weeks in structured sessions on AI in the classroom, in research, and in day-to-day academic work, then the next seven building tools of their own. For anyone tracking news articles about AI adoption in American higher education, Wentworth Institute’s program stands out because it treats faculty development as a design challenge rather than a training requirement to check off.
Want to see how a structured AI program can change teaching practice on your campus? Explore the Faculty AI Academy model and pass this article along to your academic leadership team to start the conversation.
Why Awareness Alone Fails as Faculty Development
Most AI initiatives at colleges and universities follow a predictable pattern: a keynote, a webinar series, maybe a policy statement about academic integrity. These activities build shared vocabulary, but they rarely change what happens in a classroom on a Tuesday morning. The gap between awareness and application stays wide, and it widens further when faculty are left to close it on their own time.
Larson’s listening tour confirmed what much of the recent coverage of AI at universities across the USA suggests. Faculty interest is high; practical confidence is low. Instructors do not need another explanation of what large language models are. They need protected time to experiment, expert guidance when they get stuck, and colleagues working through the same problems beside them. The Faculty AI Academy was built around those three needs from the start.
Inside the Academy’s Two-Phase Structure
Larson designed the program in partnership with Hariharan Naganathan, associate professor of Construction Management and Wentworth’s Faculty Fellow for Teaching and Learning Innovation. Naganathan taught alongside Larson and helped shape how the 14 weeks were organized. The pairing matters: one brought institutional technology expertise, the other brought the perspective of a working instructor, and the result was a curriculum that carried faculty ownership from day one.
“I wanted it designed by and for faculty,” Larson said. That principle shaped everything from the session topics to the decision to dedicate half the program to independent building.
Phase One: Seven Weeks of Structured Learning
The first half of the Academy covered AI applications in three areas: teaching, research, and everyday academic work such as course preparation and feedback. Sessions were live and interactive, and the format pushed participants to connect each topic to their own courses rather than absorb concepts in the abstract. Naganathan observed a shift almost immediately. “The conversations during the live sessions quickly moved beyond simply using AI to generate content,” he said. “Faculty began looking at real challenges their students face in their courses and thinking critically about how AI could help address some of those challenges and improve the learning experience.”
Phase Two: Seven Weeks of Building
The second half flipped the format. Faculty stepped away from structured sessions and committed to building something of their own, with support available when needed. This is the component most institutional programs skip, and it is arguably the one that made the Academy work. Time is the scarcest resource in academic life, and the program’s core promise was precisely that: time.
“The Academy serves as a platform for faculty to learn about AI and identify opportunities to apply it within their own teaching, research, and professional work,” Naganathan said.
What the First Cohort Built
The projects emerging from the pilot show how quickly stubborn classroom problems become solvable when instructors get real building time:
- Lab report support agents. Two faculty members are building AI agents that help students work through lab reports, giving students a responsive resource without replacing the feedback instructors provide.
- Handwritten equation conversion. Another participant is developing an agent that converts handwritten equations into digital format, addressing an accessibility problem in math courses that has proven difficult to solve at scale.
- Deployed engineering tools. One School of Engineering participant put a project to work in the final weeks of the summer semester, testing it with real students rather than waiting for a perfect version.
The full cohort will present these projects at the Faculty AI Academy Showcase on Oct. 8, giving the wider campus community a concrete look at what disciplined experimentation produces.
Already experimenting with AI tools in your own courses? Share your experiences in the comments below — your project could give another department a head start.
The Case for Multidisciplinary Cohorts
The pilot group was intentionally multidisciplinary. Philosophy, chemistry, interior design, mechanical engineering, applied mathematics, computer science, and construction management were all represented, and Larson credits that mix for the quality of the discussions. Faculty from different schools kept discovering they were wrestling with identical problems: how to assess work fairly, how to keep students engaged, how to give timely feedback at scale.
“We talk about silos all the time,” Larson said. “This was an opportunity to get people across campus in the same room, or at least talking about the same issues.”
For institutions planning their own programs, the lesson is clear. Homogeneous cohorts reinforce existing assumptions. Mixed cohorts surface shared problems — and shared solutions — that no single department would identify on its own.
Confronting Bias, Privacy, and Academic Integrity Head-On
The Academy did not sidestep the harder questions surrounding AI. One full module addressed bias, privacy, authorship, and academic integrity, working through classroom scenarios that participants are likely to encounter in practice rather than hypothetical edge cases. Another asked faculty to make deliberate decisions about when to encourage AI use in a course, when to limit it, and when to exclude it entirely.
This matters because policy imposed from above rarely fits the realities of individual courses. A capstone design project and an introductory writing seminar raise different integrity concerns, and faculty need frameworks they can adapt, not one-size-fits-all rules.
Bringing Students into the Conversation
ATLAS also brought students into the room for a moderated conversation with the cohort, a session participants rated as one of the most valuable parts of the entire program. Students use AI tools, misuse them, and worry about them in ways instructors sometimes misjudge. Hearing that perspective directly, before finalizing course policies, gave faculty information no survey could replicate. The student conversation returns in the fall version of the Academy — a sign of how central it became.
Five Lessons for Institutions Building Their Own AI Programs
Wentworth Institute’s experience yields transferable guidance for any university in the USA weighing a similar investment in faculty development:
- Start with listening. Larson spent months talking to faculty before designing anything. The program answered a documented need, which explains why demand followed.
- Design by and for faculty. A faculty fellow co-taught and co-shaped the program. Peer credibility does more for participation than administrative mandates.
- Give time, not just training. Half the program was dedicated building time. Without it, good intentions dissolve under teaching loads.
- Mix disciplines deliberately. Cross-school cohorts break silos and reveal problems — and solutions — that departments miss in isolation.
- Iterate on feedback fast. Larson collected anonymous feedback and rebuilt the program within a single semester. Responsiveness builds trust for future cohorts.
If you are an administrator planning professional development for the coming academic year, consider carving out protected building time instead of scheduling another demonstration session.
A Shorter, Sharper Fall Cohort
Feedback from the pilot shaped a revised fall version that runs six weeks instead of 14, with more hands-on lab time and more examples from faculty who have already completed the work. The format change reflects a mature understanding of adult learning: the first cohort needed an extended runway to establish confidence, while the second can move faster because working examples now exist across campus.
Registration for the fall cohort closed on Friday, Sept. 11, and space was limited. Faculty who could not make the scheduled times were encouraged to register anyway and flag their availability, because demonstrated demand will determine whether ATLAS adds evening or alternate sessions in the future.
Planning to apply to a future cohort? Submit your application as soon as registration opens. Limited seats fill quickly, and early interest helps organizers justify additional session times.
Advice for Faculty Who Still Hesitate
Naganathan’s pitch to skeptical colleagues is direct. “You do not need to be a technology expert to do this,” he said. “As faculty, we are constantly thinking about how to support student success, and AI can become another tool to help us do that more effectively. The Academy is really about creating a space to experiment, ask questions, and learn alongside colleagues.”
That framing — experimentation over expertise — may be the most portable lesson of the entire initiative. Faculty development programs succeed when they lower the cost of trying. Wentworth Institute of Technology lowered that cost with structure, peers, and time, and its first cohort responded by building tools that solve real problems for real students.
Have questions about how AI could fit into your discipline? Contact your institution’s teaching and learning center, or explore our related articles on AI in higher education for further reading.
The Faculty AI Academy Showcase on Oct. 8 will make the pilot’s outcomes public, and the fall cohort is already building. For anyone following higher education news, Wentworth Institute offers a case study worth watching: when faculty development gives instructors genuine time to build with AI, usable results arrive faster than most administrators expect.
