Q&A with Rachael Orellana on AI and Floodplain Management

By Quinn Fenger and Rachael Orellana

Rachael Orellana, P.E., has never been one to sit on the sidelines. With nearly two decades of flood‑risk management experience—including years spent leading project teams for the U.S. Army Corps of Engineers—she now serves as the Digital Solutions Program Manager at GEI Consultants, where she heads up the firm’s initiative on information management and numerical modeling. Rachael is co‑moderating a two‑part series at this year’s Floodplain Management Association (FMA) conference: “Panel–Part 1: Practical Applications of AI to Support Flood Risk Management” and “Part 2– From Models to Maps to Megastorms: AI in Floodplain Work.” In this conversation she shares how she wound up in the moderator’s chair, what excites her about the topic and why she hopes these sessions will get people talking.

Q: How did you end up co‑moderating these sessions about artificial intelligence at FMA?

A: It actually started last year. I went to FMA and saw Vince Geronimo from Geronimo Engineering moderate an exceptional panel on big data and AI—I mean, it was so good I downloaded ChatGPT afterward just to play with it. I found myself asking, “How can this help me personally? How can it help professionally?” When Vince invited me to co‑moderate this year, I thought, absolutely! I’m curious and I wanted to learn from national leaders in AI and digital solutions.

Q: I want to give an opportunity right away to answer the why for anyone reading; why should FMA attendees care about these panels?

A: AI is going to change our world, and it’s going to change how we manage flood risk. Last year I left Vince’s session craving more practical examples. I understood the promise of big data and AI, but I wanted to know: what are people actually doing on the ground? How is it helping day‑to‑day? That’s what these panels are about. We’re bringing in people who can share real stories about using AI for flood‑risk communication, emergency response, wetland restoration, dam safety—you name it. And I’m thrilled that Bill Kirst (AI Ambassador and Adobe’s Change Manager) is flying in from Seattle. He just published Leading Change in the Era of AI and won a CIO 100 award for his chatbot at Adobe, so he’s a big draw.

Q: That’s a good transition to my next question. Your panels feature some distinctive voices—what perspectives are you expecting them to share?

A: We ended up with a really eclectic group. Shima Shamkhali Chenar from the California Department of Water Resources will highlight floodplain management examples and, fun fact, she used to do research focused on utilizing AI to track the spread of norovirus in oysters before coming to DWR. Bill Kirst will talk about leading change during this AI transition—he brings a human touch to technology conversations. Ryan Ripken from GEI is our machine‑learning specialist; he’s been developing tools for Delta emergency response, Michigan dam safety and more. And Julia Kumari Drapkin, who founded the community platform ISeeChange, will explain how they turn community photos and social‑media posts into quantitative data. I first met her at an FMA session last year and thought, “This is really cool—communities reporting flood events and AI translating that into something engineers can use.” I can’t wait to hear from all of them.

Q: AI is already showing up in some surprising ways within the AEC world. Can you share any examples you’ve come across?

A: Julia’s examples blew me away. ISeeChange scrapes photos, Facebook posts—anything people share publicly about flooding—and uses AI to translate that qualitative data into actionable insights. One example she gave was that someone reported two feet of water on their street. They combined that with topographic data and realized there must be two feet of flooding around the corner, even though no one reported it. They checked and sure enough, that area had flooded. Being able to ground‑truth unreported information like that and put it in a format agencies can use is amazing. It’s empowering communities and helping them secure funding for flood‑risk initiatives. And I love that the platform can rearrange data to match each agency’s system—no more “our database doesn’t look like yours.”

Q: Your sessions are called “Practical Applications” and “From Models to Maps to Megastorms: AI in Floodplain Work.” What’s the difference between Part 1 and Part 2, and what can we expect?

A: Part 1—which I’m moderating—features a mix of panelists: Adobe, DWR, a private consultant and a Community Climate Data Innovator. Part 2, moderated by Vince, leans heavily on public agencies: Esri, the Army Corps and DWR. Both panels will have their own Q&A, and then we’re going to bring all the panelists up for a joint conversation during the last ten minutes. Vince is known for adding fun elements—he actually asked ChatGPT to write duet lyrics for us to sing as an opener! He sent me a chat transcript where GPT suggested custom lyrics to popular songs, making them explain AI and Machine Learning. We’re still working on it, but we’re thinking about synthesizing the sessions’ key takeaways into a song everyone can enjoy at the end. It’s going to be, as Vince says, a sensational set of sessions!

Q: The panels emphasize that AI isn’t a threat—but ignoring it might be. What concerns are you hearing, and have you heard anything from the panelists that address them?

A: People have serious questions, and they should. We need policy guardrails—we don’t have them yet. Bill Kirst phrased it perfectly during our planning call: the question isn’t “Can you trust AI?” It’s “Can you trust yourself with AI?” You need to know yourself when using these tools. There are also environmental concerns—these models require power. There are concerns about AI biases and how AI will impact human behavior. Additionally, the ability of AI to mimic voices and styles so effectively also means misinformation can spread fast. In our panels we’ll stress transparency: be open about when and how you’re using AI. And we’re not glossing over risks. We want to start a conversation about how to use AI responsibly, not just cheerlead for the technology.

Q: Any last words for readers gearing up for the conference?

A: Come early! The room only holds 100 people and I’m hoping for standing‑room only. I actually sat on the floor at Vince’s panel last year because it was packed, and it ended up being my highlight of my conference. If you’re attending FMA, I’d love to see you at our sessions and our booth. And if you can’t make it, reach out—I plan to write an article afterward and want to keep this conversation going. We’re even talking about organizing an FMA training on AI next year. AI is going to shape how we manage flood risk; let’s make sure we’re part of that conversation.