The spring conference season has officially wrapped up, and what an incredible first half of the year it has been! With a little more time to breathe, I wanted to share a few personal updates and reflect on the major AI themes that are shaping our industry.
Personal Projects & Big News
I’m diving headfirst into some exciting personal projects this summer:
Finalizing My Book: After two years of hard work—a year longer than I anticipated!—my book is nearing the finish line. I’ve just received feedback from nine reviewers and am making the final edits to get it ready for its fall release. The book is all about how to become an AI administrator, using AI to create, earn, and grow both personally and professionally. More details to come soon!
Preparing for a TEDx Talk: I am incredibly honored and excited to have been accepted to give a TEDx talk this August! The theme is "The Year 2050," exploring what our life, technology, and society might look like. This is right up my alley, and I can't wait to share my futuristic-focused talk on a subject I'm so passionate about.
Key Tech Themes from the Conference Circuit
Beyond my personal projects, I’ve been thinking a lot about the patterns I observed at recent tech events. There was a fascinating split between the topics pushed by major companies and the real-world conversations happening among practitioners.
Theme 1: The Age of Agents (and the Foundational Work We're Ignoring)
As I predicted on the Super Data Science podcast last year, the dominant theme was AI Agents. Every event I attended featured agents in some capacity.
However, the presentations from SaaS and tooling companies often glossed over the immense behind-the-scenes work required to make agents effective. From my consulting work on AI strategy, I see companies that could achieve significant wins by focusing on the fundamentals first. Before you can successfully implement agents, you need:
Strategic Alignment: A clear connection between your business goals and your technology roadmap.
Modern Systems: Upgrading the underlying infrastructure.
Digital Foundations: For some, this means simply getting off paper. It's nearly impossible to implement advanced AI if your processes are still analog.
While I believe agents are the next step in our evolution, we as developers have a lot to figure out before we get there.
Theme 2: The "Underground" Topic: Evals Are Everything
This leads to the second theme—one that wasn't on the main stage but was the central topic of conversation among practitioners: Evals (Evaluation Metrics).
Once you have a model or agent ready for production, how do you really test it? We've moved beyond simple accuracy. Today, effective evals must be specific to your business use case and might measure things like:
Latency: Is the response time for a voice model acceptable?
Brand Voice: Does the AI communicate in a way that aligns with your company's identity?
Factual Accuracy: Is the information it provides true and reliable?
The problem is, many are skipping this crucial step because creating good evals is hard. But my biggest piece of advice is simple:
Any eval is better than no eval. Just start.
Your first test won't be perfect, but you can iterate and improve it over time. Don't let the pursuit of a perfect test stop you from having one at all.
A Call for Leadership: Move from Hype to Vision
Ultimately, practitioners are excited about agents but are also realistic. We know they won't solve all our problems overnight.
This highlights the biggest gap I saw this year: leadership. Too many leaders were caught up in the hype, focusing on what they were adopting ("we're using GenAI!") instead of painting a picture of the outcome.
My advice to leaders is this:
Don't just tell me you're adopting AI. Show me what a better future looks like for your organization because of it. Go back to being the visionary. A clear vision of success is the most important component needed to navigate the AI revolution successfully.
Your Turn
What were the hits and misses you noticed this conference season?
And more importantly… what are your summer plans?
Let’s make this a two-way conversation. Drop a comment below — I’d love to hear from you.
xoxo,
Sadie
A focus on the how and what is not as powerful as the why. Just implementing AI to be able to say we did it will not improve a business. Some type of outcome needs to be aimed at or the “why” for implementing it that is more then “ so we can say we did”.