AI for All: Empowering Non-Coders to Build Real Solutions

Have you ever encountered a task or process that you believe could be completed more efficiently with the help of an AI-powered tool? Perhaps there's a small pain point in your workflow, but the idea of creating an AI-aided solution sounds overwhelming.

The Citizen Developer Workshop, hosted by Yale School of Medicine's Educational Technology & Innovation (ETI) team, was designed with exactly these challenges in mind.

“Citizen developers” are non-IT professionals who use approved, no-code platforms to build practical applications on a small scale. The goal is to empower users at the point of need, reduce bottlenecks, and remain compliant with Yale’s privacy and security standards.

“At Yale School of Medicine, we are fortunate to have enterprise-level access to a variety of AI platforms,” said Jaideep Talwalkar, MD, associate dean for educational technology and innovation. “These platforms offer all of us the ability to build solutions while offsetting data privacy concerns present in direct-to-consumer products.”

Over the last year, the ETI team has hosted a series of technology-focused events ranging from ideathons to panel discussions. However, this is their first event focused entirely on empowering non-coders to experiment, build tools and applications, and launch solutions to technical challenges.

“The people closest to the work often have the best ideas for improving it,” noted Kathleen Ludewig, PMP, co-lead of educational technology & innovation.

The half-day workshop, led by Talwalkar and Ludewig, featured mini-teach lessons and demonstrations by ETI team members Gary Leydon, associate director of medical education technology, Ye Xu, PhD, business analyst lead, and Vincent Randolph, business systems analyst & project manager intern.

The short presentations gave attendees an overview of what tools are available at Yale, which tool is best suited for a specific project, and how to start experimenting.

“The workshop was a great chance to see what folks across YSM are building with Yale’s AI tools and to realize that these platforms are far easier than they look once you get your hands on them,” said Nickia Corley, administrative director in the Office of Global Health Education. “ETI workshops are the kind of session where you show up curious, learn what’s possible, and walk out thinking, ‘Oh, I can actually do this.’”

Randolph covered the basics for setting up a knowledge base and choosing the right AI tool for your needs. He gave attendees a series of questions to consider when forming a knowledge base, such as “do I have copyright permission to use this source” and “is the material current and complete enough for the task.” To develop a strong prompt, Randolph suggested using the CRAFT (context, role, action, format, target) model to ensure the AI has all the information needed to deliver an accurate response.

Randolph then explored selecting the right tool for a project. Custom chatbots, for instance, are best used when your desired output is guidance or feedback. They can help refine drafts or support decisions but remain “in the conversation,” meaning they cannot act outside of a question-and-answer format. In contrast, a custom agent can execute more complicated workflows and multi-step tasks.

Leydon led attendees through a demonstration of customization within Yale’s AI platform, Clarity.

Customization lets users build an AI chatbot for their needs by limiting the knowledge base to only what they upload. Users can choose an AI model available in Clarity (GPT, Claude Sonnet, etc.) and adjust the system prompt with detailed instructions to get accurate results and the desired output.

Clarity also allows users to collaborate on customizations with other Yale members, offering three roles to facilitate this process (owner, contributor, and reader). A benefit of developing within Clarity is that security measures and data compliance are built in, meaning users have a safe environment in which to experiment and explore.

Xu gave a demonstration on Google’s NotebookLM, another AI product available to users at Yale.

NotebookLM relies on a user-uploaded set of materials to create outputs. Showcasing the versatility of the platform, Xu illustrated the range of outputs offered including quizzes, infographics, and even slide decks based on the source materials.

Armed with knowledge, tools, and encouragement, attendees explored AI solutions as members of the ETI team offered guidance and feedback.

Participants were encouraged to share their creations with attendees. Tiffany Hu, MD, MSCR, demonstrated how she used Gemini Canvas to create a clinical escape room-themed game for internal medicine residents.

“The experience wasn’t simply about learning AI—it was about understanding how to use different tools strategically and turn ideas into practical solutions,” said Meenakshi Alreja, PhD, associate professor of psychiatry and neuroscience. “In a single afternoon, I uploaded one of my lecture presentations into NotebookLM and watched it generate quiz questions, create an engaging audio summary, a mind map, and suggest ways to improve my teaching. Some features were remarkably effective, while others highlighted the current limitations of AI. The workshop reinforced that the future isn’t about finding one perfect AI tool—it’s about knowing which tool is best suited for each task.”

Karen Angelicola, project coordinator in the Office of Student Research, found the day’s activities both valuable and inspiring. "The workshop gave me a much better understanding of the AI tools available at Yale and sparked ideas for improving student support. I'm now exploring whether we can build an AI assistant that helps medical students navigate our thesis guide and quickly find answers to common questions."

Second-year MD student Lilianne Sutton enjoyed the opportunity to connect with faculty and staff and share her perspective on AI-powered tools in medical education.

“While I didn’t make much progress on my original idea, I experimented with several different platforms and am pleasantly surprised by the range of tools we have access to through Yale! I now have a really long list of ideas I’ll be tinkering on, from creating a simple website as a med ed hub for rare diseases to a more AI-powered tutor for a specific clinical skill.”