Citizen Developers Explore New Enterprise Access to Claude

Citizen developers, non-IT professionals who use no-code artificial intelligence (AI) platforms to build solutions, are on the rise across Yale School of Medicine (YSM). Yale’s campus-wide access to AI tools has empowered faculty, staff, and students to explore creative solutions to their workflows.

Indeed, many users are becoming more comfortable with these tools thanks to events like the Citizen Developer Workshop hosted by the Educational Technology & Innovation (ETI) team. The half-day workshop gives both experienced AI users and the AI-curious an opportunity to sit down with an IT professional, work through a problem, identify an AI-augmented strategy, and start developing right away.

“Through these workshops, we aim to equip people with knowledge and skills to help shape how AI is integrated into their work,” said Jaideep Talwalkar, MD, associate dean for educational technology and innovation. “By exploring Yale’s enterprise AI tools, we collectively discover ways to strengthen our work and amplify our contributions.”

The most recent workshop, which took place on Oct. 16 in the Medical Library’s team-based learning classroom, gave attendees the chance to learn about, and experiment with, Yale’s new enterprise-level access to Claude by Anthropic.

Announced on Sept. 15, the enterprise agreement allows faculty and staff to purchase a Claude license for access to Claude Chat, Code, and Cowork. Notably, people in the medical school or other units that work with electronic protected health information (ePHI) are automatically given access to a HIPAA-compliant environment within Claude Chat and Code with their enterprise license.

During the workshop, IT Director and Co-Lead of Educational Technology & Innovation Kathleen Ludewig, MPP, MSI, walked attendees through a comparison table explaining the differences between personal and enterprise Claude environments as well as the license options available for faculty and staff.

To illustrate Claude Enterprise’s capacity, Chelsea Wallace, MEd, manager of pre-clerkship courses and longitudinal coaching, shared her experience using the tool.

Wallace, a self-identified AI novice, oversees 14 courses in the MD curriculum. At the end of each course, she compiles summaries of student evaluations, which can then be shared with course directors. Creating these reports takes considerable time because it involves combining large amounts of data and manually preparing each document.

Recognizing the potential for an AI-assisted workflow, Wallace enlisted the help of Claude Enterprise’s Skills module to create a set of instructions and rules for creating evaluation reports from raw data. Skills enable users to save, refine, and reuse instructions for specific workflows, reducing the time and effort needed to complete recurring tasks.

After developing and refining the skill, Wallace was able to create one-page evaluation summaries for each course director within minutes. Claude also synthesized previous years’ evaluations and added charts showing course satisfaction over time.

Wallace said the generated reports were not entirely free from error, and she emphasized that human review was needed to ensure the summaries were accurate and appropriate. However, the workflow showed promise.

“I still consider myself a novice when it comes to AI, so sharing my experience at the workshop was a great reminder that you don’t have to be an expert to share what you’ve learned and help others,” said Wallace. “That’s what I appreciated most: we were all learning from one another, regardless of experience.”

Attendees also heard from Joseph Canterino, MD, MHS, associate professor of medicine, who attended June’s Citizen Developer Workshop and was inspired to build a video game. Canterino used ChatGPT to create Viral Invaders, a clinical reasoning adventure game designed to help residents develop problem-solving skills.

Following the peer demonstrations, Ye Xu, PhD, business systems analyst, and Gary Leydon, associate director of medical education technology, highlighted many of the features available in Claude Enterprise, including Projects, Skills, and Claude Code. Leydon also gave demos of custom Skills and how to search a library of pre-built Skills that may suit a user’s needs.

Attendees were invited to begin experimenting with Yale's suite of AI tools and building solutions as members of the YSM ETI team circulated and offered guidance. Ludewig also shared a Citizen Developer Toolkit containing links to relevant resources and tutorials that users can consult after the workshop.

"I learned a lot about our options for AI platforms at Yale and what they can do for us," said France Galerneau, MD, professor of obstetrics, gynecology and reproductive sciences. "The workshop was a wonderful opportunity for hands on work on a project with guidance from the faculty. I also enjoyed hearing about work that others have done using AI in medical education."

At the end of the workshop, three attendees shared demos of what they created.

The projects included three-dimensional animations of echocardiograms using Claude, patient handouts on pulmonology using Claude Enterprise, and R trend analysis of infectious disease data using Clarity AI.

“One of the things that struck me about the participant demos, without any prodding from the YSM ETI team, is how each participant spoke about having validated educational content as input for AI and having expert review of the output as well,” said Ludewig. “Several of them also mentioned they chose AI for the skills they didn’t have, such as graphic creation or coding.”

The opportunity to build products and solutions using tools that extend one’s expertise is exactly the goal of Citizen Developer Workshops. The YSM ETI team hopes each attendee feels confident to explore these tools going forward.

For Galerneau, it's done just that. "I am glad to have met the ETI team and have already reached out for help since the workshop took place."

Keep an eye out for future Citizen Developer Workshops as part of the YSM ETI TIME series.