Case Study: My Own AI Journey

Broward County Public Schools, 2024 - 2026

Where I Started

My relationship with AI did not begin with a single moment of discovery. It grew gradually, alongside years of professional development. As a Microsoft Innovative Educator (MIE) Expert since 2021, I recertify annually to stay current, and when AI components began appearing in the MIE Expert curriculum, I was ready. I had been completing courses on Microsoft Learn, attending webinars, and independently researching AI tools ranging from Gemini and ChatGPT to Claude, Microsoft Copilot, and Apple Intelligence, building my knowledge across platforms long before AI became a formal priority in my district.

When our department was tasked with leading the district's AI rollout, I leaned further into that learning. I began using Microsoft Copilot more intentionally, experimenting with prompt engineering, and thinking critically about what AI could realistically do for educators and students. I was already building the foundation I would need to lead others.

Immersing Myself in the Field

In December 2024, I attended two days of training at the Microsoft office in Fort Lauderdale, AI Fundamentals on December 11th and Azure AI on December 12th. That same month, the district's AI Task Force launched with our department at the helm. The AI Task Force is a district wide group with representatives from every department, built on a framework developed by my director. I became an active member from the start, joining a team focused on reviewing, editing, and refining the documents that would guide the district's ethical and practical approach to AI. This work included contributing to the Ethical Implementation Guide and a series of one pagers developed for teachers, students, parents and guardians, and district staff and leadership. The guide was distributed to the district in draft form with the acknowledgment that it would remain a fluid document as AI continues to evolve rapidly. My director led the majority of presentations to the task force, I presented on several occasions, and other colleagues contributed to and presented additional sessions throughout the process. I attended nearly every meeting. The AI Task Force paused after April 2026 and is expected to continue in the fall.

In January 2025, our department along with the other members of the AI Task Force began a two-month Copilot premium pilot, giving us hands-on experience with the tool in a real context before it rolled out more broadly. That same month I attended the Future of Education Technology Conference in Florida, where I participated in sessions focused on AI across multiple platforms including Apple Intelligence, which was particularly relevant given my work with our VILS schools and their iPads.

From February 27 through March 1, 2025, I attended the AI K12 Deeper Learning Summit in Anaheim alongside my director and one of my colleagues. One of the most significant takeaways from that conference was how students and teachers are already leveraging AI in their daily lives, with or without guidance, and how urgent the need for AI literacy truly is. That reality reinforced the importance of equity and impact with digital tools.

In May 2025, I completed a series of Microsoft Learn courses and earned badges and trophies in AI for Educators, Introduction to Generative AI for Trainers, and several modules focused on prompt engineering, Copilot for teaching and learning, and supporting learners with AI tools. On May 28th and 29th, I organized and participated in a formal two-day Microsoft training for our entire department, focused on Copilot 365, agents, and practical use cases in M365 productivity tools like Outlook and Excel. Much of what was covered on Day 1 solidified knowledge I had already built independently. The most significant new learning came in the areas of agent building and Copilot's capabilities within Excel, which expanded how I thought about what the tool could do beyond basic prompting.

The start of the 2025-2026 school year brought a new wave of AI momentum. Over the summer, principals across the district selected site-based AI champions at their schools, known as AI Liaisons. Our department kicked off the year with additional Microsoft training, and our AI Liaison cohort lead began building out a series of learning experiences for the AI Liaisons covering a variety of tools including MagicSchool, using Copilot to build materials for Promethean boards, and creating agents. Members of our department attended the AI Liaison meetings to ensure we were properly prepared to provide support. The AI Task Force also started up again, and it was during this phase that our collaboration with NCCE deepened, with their input informing the documents our task force was reviewing and refining.

In January 2026, I completed a formal Microsoft Copilot Studio training run by a Microsoft partner, alongside three ITFs and our AI Liaison cohort lead. As part of the training, we built an agent in our own environment, and I received a copy of Microsoft Copilot Studio Step by Step: Customizing Copilot and Creating Agents as a reference resource for continued learning afterward. That same month I joined the Student Facing AI Testing Group, a district level team responsible for formally vetting AI tools for potential student use. The group used a rubric to evaluate tools consistently across evaluators. When the rubric was presented for feedback, I recommended replacing vague qualitative descriptors such as the word "some" with numerical values to improve reliability and consistency across reviewers. That recommendation was incorporated into the final rubric.

In March 2026, I completed a series of MagicSchool certifications including Foundation AI, Intermediate AI, Magic School for Students, Adobe Image Generator, and Level 1 and 2 for Schools and Districts Org Admin courses. While MagicSchool is primarily an education focused platform, completing these certifications reflected my commitment to understanding AI tools across multiple ecosystems and staying current with what educators are actually using in their practice.

Contributing to District AI Strategy

Beginning July 28, 2025, I participated in weekly Copilot Phase 1 implementation meetings that ran through December 15, 2025. These meetings were focused on the practical rollout of Copilot across the district, tracking progress, troubleshooting issues, and ensuring alignment across the district.

On September 18, 2025, I attended the first NCCE meeting, and beginning October 20, 2025 those meetings became a weekly standing commitment every Monday. In those meetings, which include my director, several IT directors, and the Executive Director of IT, I contribute to discussions about upcoming trainings, district AI guidelines, and Copilot usage data. I take notes and follow up on action items to keep the work moving forward between sessions. It was part of this collaboration as our team began bringing NCCE feedback directly to the AI Task Force, creating a feedback loop between district leadership and the task force's document refinement work.

Leadership Presentations

On June 9, 2025, one of my Instructional Technology Facilitators presented a Google NotebookLM session at Broward Leadership Days for district administrators, and I was present to field any district specific questions. In preparation for a potential broader Google rollout, I also built a Canvas course for the NotebookLM training. While the course was ultimately not used this year because the decision was ultimately made to continue Microsoft Copilot training first, building it proactively reflected my approach to staying ahead of where the work might go.

On June 10, 2025, I coordinated bringing Microsoft vendors to Broward Leadership Days to deliver a Copilot training session for assistant principals and principals at an event attended by approximately 1,000 administrators. I was present throughout the session to field district specific questions that required local context and knowledge.

In November 2025, I presented to approximately 20 district directors at a leadership meeting on the topic of creating agents. My director opened the session, and I led the presentation on agent creation. As part of the live demo, I built a Leadership Feedback Coach agent in real time, using documents provided by one of the directors who had identified repetitive feedback tasks she wanted to automate. My colleagues joined in as needed throughout the session.

As the school year progressed, I organized many other opportunities for the staff and a variety of groups to be trained directly by vendors, making sure that there was continuous learning happening for staff so we could be ready to support the teachers.

Building Capacity Across the District

As my fluency with AI grew, so did my role in helping others build theirs. I have become an informal AI resource across the district, with teachers and staff forwarding questions directly to me for more detailed guidance on prompt engineering, Copilot use, and other AI tools. Questions range from basic how-to support to more nuanced conversations about policy, instructional application, and responsible use. I respond individually, meeting people where they are in their own AI journey.

When one of my ITFs created an exceptional AI job aid, I recognized its value immediately and worked to incorporate it into our department's district wide technology Canvas resource, Digital Resources, Easy as 1,2,3, making it available to all staff across the district. I am also currently involved in the development of a district wide AI resource that will provide staff with broader access to guidance and support at scale.

I have also presented AI Best Practices to assistant principals at the district wide Assistant Principal Meeting, guiding administrators on AI adoption, instructional impact, and policy considerations to help school leaders make informed decisions about technology use.

Beyond the district, I model intentional AI use at home, teaching my own children how to use these tools responsibly and what AI literacy means in practice.

Building and Experimenting with Agents

One of the areas where my AI learning has gone deepest is agent building. I have created a variety of agents for both professional and personal use, each one teaching me something new about how these tools work and where they fall short.

Professionally, I built a technology integration agent designed so that a teacher could describe an instructional need and receive a suggested tool in response, along with a prompt to schedule time with their Instructional Technology Facilitator for implementation support. The concept was to use AI as a triage layer that connected teachers to human expertise rather than replacing it.

I also built an onboarding buddy agent loaded with department onboarding information, allowing new staff to ask questions and get answers independently. The concept directly extended the onboarding work I had been doing since 2023, this time using AI to make that information conversational and accessible on demand.

I attempted a department navigator agent as well, though it remains a work in progress. The challenge has been source material. With organizational changes happening frequently and outdated information still accessible on the district website, keeping the agent focused on accurate and current information has proven difficult. I continue to revisit it when I can, and the experience has deepened my understanding of how critical clean, consistent source material is to effective agent design.

The Leadership Feedback Coach, built live during a director presentation in November 2025, automated a repetitive feedback task identified by one of the directors in the room. Using documents she provided on the spot, I built and demonstrated the agent in real time.

On the personal side, I built a Gemini gem for my wardrobe. I uploaded a detailed spreadsheet of my clothing into the gem and now use it to generate outfit suggestions based on the event I am dressing for. It is a small thing, but it is a perfect example of how I think about AI. Find a real problem, build something practical, and see what happens.

Where I Am Now

I am a lifelong learner. I recently completed HP's AI for Business Professionals course and continue to pursue courses and resources that deepen both my AI knowledge and my understanding of the business contexts where that knowledge applies. Learning is something I do because the field moves fast and staying current is part of doing the work well.

Today I use AI intentionally and purposefully, most often as a thought partner and final reviewer. I draft everything myself first. Then I use AI to give it one more look. I find it especially valuable for the kind of work that causes human fatigue, reviewing the same document multiple times, catching what fresh eyes would catch. Unlike a human reviewer, it never gets tired of looking at the same thing twice.

I do not use AI for everything. I use it where it adds genuine value, and I am always thinking about how to help others do the same. That is what drives my work, not the technology itself, but what it makes possible for the people using it.