I’ve designed a thousand features over my career and tested hundreds of tools to help me be faster at my work. Let’s break it down by each stage of the design process.
Stage 1: Research / PRD
Before jumping into pushing pixels, most of us collect our thoughts in product requirement documents, where, in the case of a feature, we structure thoughts, requirements, goals, and other important pieces that help us not get stuck while actually designing UI.
I used a few tools that speed up this stage, but by far the best ones are ChatPRD. It's an AI tool that helps you draft those product documents or clean up your handwritten thoughts into a cohesive, structured plan.
It also comes in handy to be prepared before you jump into AI UI tools. The quality of your prompt and thoughts makes a big difference.
Stage 2: Idea to UI
Traditionally, designing UI meant going to Figma and spending hours looking at the blank canvas before coming up with something.
With new tools like Banani AI, you can just describe the screen you want in plain text as a text prompt, and they will make drafts of UIs and wireframes.
As I wrote in the previous section, the better your thoughts are outlined, the better AI UI results will be, so please, don’t neglect that step.
Once you’re happy with the first screen you generated from the prompt, you can simply click on "interaction" elements from the user perspective, and Banani will try to predict what should be on the next screen.
It can also apply your color palette or design system of your product to design UIs that match your product’s look and feel.
This is a big deal for teams that need consistency across multiple products and don’t want to manually re-style every AI-generated mockup.
Stage 3: Iteration and feedback
Design is never right on the first try. Depending on your workflow, you probably will pass design files around, gather feedback, and redo screens manually. With AI, iteration feels more like a conversation.
Tools like Granola that are intended for meeting transcriptions can greatly simplify the aggregation of the feedback from design critique calls, user testing sessions with real people, or just a friendly call with your design team mate.
Additionally, there are now tools that help you run uncontrolled ux testing and get all of the same benefits. One that I use is Maze. It lets you send a link to the design to the user, and for them to go through the flow.
You can use aggregated summaries and send them to UI generative tools from the previous section to make the next iteration.
Tl;DR
Long early stages of research, layout tweaking, styling, and iteration soon will be a distant memory. Adopt new AI products and design apps like ChatPRD or Banani by your side to speed up the long or boring parts of the UI design process.
