Becoming Agentic

Stay curious. I believe curiosity is the beginning of every meaningful breakthrough. — Marvette Johnson

In December 2025, I attended what would be my last All-HR meeting at Webster Bank. On the agenda was a fireside chat with the CEO, the COO, and the CHRO.

Along with getting to know our leaders better through a few fun-fact questions, we wanted insight, as an HR organization, into how we could better align our department and personal goals to meet the Bank's 2026 strategic priorities.

As the fireside chat wrapped up, each leader shared a few words of wisdom that answered our overarching question. To paraphrase: 1) the times, they are a-changin', and 2) we (as individuals or as an organization) cannot use yesterday's skills and expect to succeed in an age of rapid technological growth.

Well, that was mid-December.

Over the next few weeks, those words kept ringing in my ears.

I think it was because I had been feeling it myself: "Marvette, you spend most of your professional time skilling and upskilling others. When are you going to do that for yourself?"

In January 2026, an opportunity presented itself that I simply had to take. It allowed me to step away from my professional responsibilities for a while to 1) immerse myself in emerging technologies, AI in particular, and 2) brush up on my core L&D skills (adult learning, organizational development, and project management, for example).

“Marvette, you spend most of your professional time skilling and upskilling others. When are you going to do that for yourself?”

First up: learning more about AI.

Why start with AI?

Like many people who aren't in tech or technical roles, I was familiar enough with AI. I mostly used it as an everyday advisor and productivity partner for personal, non-work tasks. As a leader of learning and development teams who wanted to be a better strategic partner, though, that was no longer enough. I needed a deeper understanding of this emerging, general-purpose technology, which experts predict will trigger a "tectonic shift" in Learning and Development (L&D) over the next two years.

I didn't believe this would require me to become a data scientist, an ML engineer, or an AI architect (thank goodness!). I did believe it would require me to learn more than AI history, terminology, and how to write better prompts. That information is easy to find.

To scale enterprise-wide training and prepare my team for this shift, I needed more than a surface-level understanding of AI. I needed to understand the infrastructure underneath it: how these systems are built, how they work, and ultimately how they can be deployed and managed.

That understanding would let me think more strategically about how this technology could align with organizational goals, and how our learning strategy would need to evolve alongside increasingly capable AI systems.

Where I started

When I began this journey, my understanding of artificial intelligence was pretty much limited to my experience as a user.

I knew what I now recognize as traditional AI: the systems most of us interact with every day, often without thinking of them as AI.

• Google search. Enter a question or a few keywords, and the system retrieves information relevant to what you're looking for.

• Recommendation engines. Shop on Amazon or another e-commerce site, and based on what you've searched for, viewed, or purchased, the system recommends what you might want next.

And then there was generative AI. I knew chatbots. I'd used them. I understood, at least at a basic level, what I could do with them. But that was pretty much it.

I knew AI from the user's perspective. I didn't really understand what was happening underneath: the models, the architecture, how information was retrieved, how systems reasoned or made decisions, how tools could be incorporated, or what made an AI system capable of doing more than generating a response.

That distinction, between using AI and understanding AI, became the reason I wanted to go deeper. I wanted to understand what was under the hood.

When I first came across the term agentic AI, I didn't have a clear mental model for what it meant. I understood AI as systems that could retrieve information, make recommendations, or generate responses. I hadn't yet started to think about AI systems that could reason, plan, use tools, make decisions, and take action toward a goal.

I knew there was something I didn't understand. I just didn't know yet how much there was to learn.

Why Johns Hopkins' Agentic AI program?

So my search began for the right program. I wanted one that would give me deep technical intuition, with a specific focus on agentic AI and AI infrastructure, without requiring me to become a software engineer or developer, or to work as a practitioner.

The market seemed flooded with generic generative AI courses. Many top business schools offered AI programs, but those focused almost entirely on high-level corporate governance and skipped the technological infrastructure. Only a handful of programs took the structured, under-the-hood approach I was looking for. Of those, the Johns Hopkins program stood out.

After reading the program brochure, I felt the program was built for someone like me: a seasoned, non-technical professional looking to move beyond basic AI literacy and achieve true AI fluency.

Here's the part of the brochure that sealed the deal:

The Certificate Program in Agentic AI is a 16-week online program designed to equip professionals with the skills to build autonomous, goal-driven AI agents capable of perceiving, reasoning, and acting independently. Combining foundational theory with hands-on Python projects, this program offers a deep dive into designing intelligent systems that can plan, learn, and collaborate to solve complex challenges.

Learners will explore agent architecture, symbolic reasoning, Belief-Desire-Intention (BDI) models, and modern techniques that incorporate Large Language Models (LLMs). The curriculum also includes reinforcement learning, multi-agent systems, and human-agent collaboration frameworks. Ethical considerations in building autonomous systems are an essential part of the learning journey.

The curriculum design lined up with my goal of preparing my enterprise for autonomous operations:

• Sixteen weeks signaled a rigorous, immersive curriculum rather than a superficial overview. That timeline promised the depth needed to truly understand complex architectural principles.

• It went well beyond prompting. The syllabus targeted the exact technical mechanics I was looking for and promised to demystify how autonomous systems are built, deployed, and managed.

• It matched my mission as an L&D leader. The explicit focus on multi-agent systems and human-agent collaboration frameworks provided the theoretical foundation I needed to picture a future where people work alongside goal-driven, reasoning machines.

• It took ethics seriously. Because my focus is enterprise-wide, the program's emphasis on the ethical considerations of building autonomous systems met my requirement for strong technical governance and risk alignment.

Ultimately, this program was the only option that promised to turn complex engineering concepts into the strategic, architectural intuition I needed to align autonomous technology with long-term workforce development goals.

Generative AI or agentic AI?

Johns Hopkins offered two certificate programs that interested me: Applied Generative AI and Agentic AI. From what I could gather:

Agentic AI it is!

Taking the plunge

The next step was a call with a program manager (the program is run by Great Learning) to learn about the program's requirements and time commitment. The call was also a chance to share my concerns as a non-tech professional and confirm that the Agentic AI course was the right choice for me.

The program manager completely understood my concerns and assured me that I "didn't need to be a programmer to take the Agentic AI course."

She even gave me an escape hatch.

I could take the first couple of weeks of the course and, if at any point I "felt like I was in over my head," transfer into the Applied Generative AI course or another Johns Hopkins certificate program.

With that escape hatch in place, I took the plunge and enrolled in the March 2026 cohort.

Follow along

And there you have it: the beginning of my journey into agentic AI. Now you know:

• What sparked my interest in this journey

• What prompted me to enroll in the Johns Hopkins Agentic AI program

• Most importantly, what I was hoping to get out of the program when I clicked "Enroll"

If you want to know more, I invite you to join me as I chronicle the 16 weeks of the program in a series of articles. These articles won't focus on the technical side of agentic AI (what LangChain, LangGraph, or RAG are, for example). Believe me, there are people much more qualified to teach you those concepts.

Instead, I'll take you along as I journey from AI literacy to AI fluency: the pre-work, the concepts that challenged me, the moments when things finally clicked, the projects and failures that taught me the most, and how my understanding grew from generative AI to agentic systems.

Here's what I'll share:

• What it was like to become a learner again. From my first encounters with Python to concepts that initially felt completely foreign.

• The moments when things began to click. The ideas and experiences that changed how I understood AI, from LLMs and RAG to reasoning, planning, agents, and intelligent systems.

• What happened when I started building. The projects, tools, workflows, experiments, failures, and breakthroughs that helped me connect the pieces.

• How my mental model of AI evolved. From thinking of AI as something that generates answers to understanding systems that can reason, plan, use tools, act, collaborate, and adapt.

• What I discovered beyond a single agent. Multi-agent systems, reinforcement learning, interaction, embodiment, and the expanding possibilities (and complexity) of AI.

• What I learned about putting AI into the real world. Deployment, operationalization, evaluation, governance, risk, and the responsibilities that come with increasingly autonomous systems.

• What changed for me personally, professionally, and as an L&D leader thinking about the future of work, learning, and leadership.

• Where I'm going next, and why this journey doesn't end with Johns Hopkins.

Most of all, I'll share what it was like to learn something that initially felt far outside my comfort zone, and what sixteen weeks of staying curious taught me about learning itself.

This isn't a technical tutorial. It's the story of learning something new, one thread at a time.

Up next in this series: Learning to Speak AI.