I came to this with no expertise in water risk, which is why I picked this subject.
I learnt about water management data sources and had a lot of fun building! My top learnings:
🌍 Data is rarely consistent across regions
OK, I know this well from my work at LandScale, but this time I had only Claude to rely on as a subject-matter expert. Different sources use different methodologies, levels of detail and geographic coverage. Creating a unified experience to answer a question like “where does my drinking water come from?” requires careful wiring and judgement.
🧑🏫 Presenting a result is not enough
The interface also needs to explain what that result means to anyone new to the subject, without oversimplifying it for those with deeper expertise. I integrated an AI explanation into the map for this purpose.
🤖 Building is one of the best ways to learn
We have an unprecedented opportunity to learn about any topic using AI tools like Claude, and building is really one of the best ways to learn. Start with a thought partner to explore a question (e.g. “who collects and shares data about drought, floods and other water-related risks?”), then a project manager to help put all your research, ideas and thoughts into place. I love using Claude design to turn my doodles into high-fidelity designs, and then of course there’s a tireless engineering team that deploys overnight or while you’re making dinner.