Intersection 007: AI-Generated Platform Prototypes
- Erin Wright
- Sep 23
- 2 min read
Updated: 1 hour ago
Ever feel like you’re working harder than you need to?
Maybe it’s hours spent building a prototype from scratch when a quick mockup would do. Or maybe it’s wrestling with broken data the night before go-live.
In this edition of The Intersection, we're sharing two things we've been talking about at ODNOS: how we’re using AI to prototype platform portals faster, and our data team's advice for managing data when the stakes are high.
AI-Generated Platform Prototypes
One of the best things about no-/low-code platforms is how fast you can spin up a mockup. Want to validate a data schema? Build it. Want to test a process? Build it.
But portals? They're trickier. A realistic mockup takes more time to configure. And if you hand it off to a designer working outside the platform's native design system, you often end up with something beautiful but impossible (or too expensive) to build.
To help us with this, we’ve been experimenting with prototypes for Salesforce Experience Cloud using v0 by Vercel, an AI UI generator that creates user interfaces from natural language prompts. With v0, we can spin up site concepts in minutes.
It’s been especially useful for showing what’s possible in native configuration, validating user flows and features, and spinning up semi-realistic mockups to explore future state solutions.
And if you want to try it yourself, you can download our prompt → here!
Your Launch Lives or Dies on Data
Digital transformations live or die on one thing: data. So why is it always overlooked?
Changing your data structure almost always surfaces the mess: duplicates, outdated categories, unclear ownership. Nikola Njegovan and our data team pulled together answers to the questions we get asked most, including how to avoid the four pitfalls we see again and again:
Partial Tests
One-and-Done
No Dedicated Sandbox
Missing Stakeholder Buy-In
They sound straightforward, but we see these corners cut far more often than you’d think. If you want a predictable and less stressful launch (because no go-live is stress free), you need to prioritize your data as much as the technical build.
Cut corners here and you risk losing something that’s hard to win back: your team’s trust.
Want to avoid making these mistakes? Read all our teams' advice → here!
Let’s Talk
It doesn't have to feel this hard.
