← Index/Case 01

Arqgen · 2026

From layers to a new foundation.

Arqgen had grown from a small proof of concept into a much broader product. The experience had grown with it. The structure was starting to struggle.

Role

Senior Product Designer

Company

Arqgen

Type

New platform

Scope

Product structure, UX, UI

Status

In development

Chapter 02

The product had outgrown its structure.

The first product was fairly simple. You uploaded a DWG file, configured the urban restrictions, defined the project parameters, generated feasibility solutions, looked at the results and downloaded what you needed.

In about a year and a half we added more than fifty features, and other solutions and projects moved into the same platform and the same interface. The product was growing in layers. A lot of those layers were useful. Together, they were getting heavy.

Some features came from very specific client needs and ended up barely used. Some of the newer tools had weak entry points, so they felt attached to the product instead of part of it.

2024

POC

2024

MVP

2025

First product

2025 — 2026

50+ features

2026

New platform

Chapter 03

A different way to experience the product.

The old experience followed a fairly linear path, built around uploading and processing a file. The new version starts to move away from that logic, opening up more ways to create, define, explore and understand a solution.

Before

UploadConfigureGenerateAnalyzeDownload
becomes

Now

CreateDefineGenerateExploreUnderstandDownload

What changed in practice

Create & Define

  • Upload a file or draw the site directly in the platform
  • Discover urban restrictions through chat or fill them manually
  • Request or fill the architectural program

Generate & Explore

  • Generate feasibility solutions
  • Explore solutions in 2D and 3D
  • Generate project quantities

Understand & Download

  • Ask questions about the project through chat
  • Download the solution

Chapter 04

The thinking came first.

The build was fast. The thinking behind it wasn't.

Early in 2026 the PM and I started looking at the platform as a whole instead of feature by feature. That conversation didn't start with the three-week build. It had been forming for months.

It came from client calls, from what Customer Success kept hearing, from using the product ourselves, from technical needs, from a fairly critical read of the existing experience, from requests clients had been making for a long time, and from a functional Figma prototype we had built earlier in the year.

  • Client conversations
  • Customer Success feedback
  • Our own product usage
  • Technical needs
  • Critical analysis of the experience
  • Long-standing client requests
  • An earlier Figma exploration

Chapter 05

Three weeks to make it real.

The new platform had to be ready to show at an industry convention, which gave us roughly three weeks to turn the thinking into something that actually worked.

A Design Sprint and a multidisciplinary workshop helped surface the gaps we had been avoiding, and gave the technical team room to work through implementation problems together.

Three solutions went into the new platform. The core Arqgen product was never in question. The other two were chosen on scope, time and how realistic the integration was. We already had one complex product to bring in, so the rest had to be simpler.

Claude helped us turn the problems and requirements into a first functional prototype. It was rough and inconsistent, but it gave us something to react to. From there I took the flows into Figma and worked through interaction, hierarchy, UI, alignment and the visual language, so the different parts of the platform would feel like one product.

AI accelerated the first draft. Design gave it structure, consistency and intent.

Previous explorationDesign Sprint / workshopPrototypeFigma refinementParallel implementationNew platform

Early exploration

Refined

Chapter 06

Designing for coherence.

Several developers were building different parts of the platform at the same time, which produced a constant stream of small questions. Should this behave this way or that way? Where does this live? What happens when there is nothing here yet?

I became the reference point for a lot of those decisions, and the Figma became the shared reference for implementation. That is what kept interaction patterns, hierarchy, UI and visual language consistent across areas that were being built in parallel.

The product decisions themselves were made closely with the PM, with the technical team involved throughout. My part was strongest in the interaction design, the UI and the refinement that made the whole thing feel like one product instead of three.

Chapter 07

The first version.

Three weeks were enough to establish a new foundation, not to finish the platform.

What exists today is a new interface, a new interaction model, a new product logic, three integrated solutions and a wider workflow that runs from creating a site through exploring solutions and downloading them.

Chapter 08

What comes next.

The platform is still being built and still needs to be validated. The longer-term direction includes more complete project organisation, a Places area that today is only a placeholder, deeper team management, more workflows and the other parts of the broader vision. None of that exists yet.

The first usability test with Customer Success happens right after this page goes up. So far the only signal we have is internal, from our CEO: "The new UI makes me want to keep generating other solutions." It is a nice thing to hear, and it is not a result. It still has to be tested with the people who use the product every day.

The platform is taking shape. The next step is finding out how well the new structure holds in people's hands.

* No confidential information disclosed. All details from public marketing materials and social media.

Next project · 02

ArqgenLand Feasibility Platform

Always happy to have a good conversation!
Feel free to say hi :)

© Nathalia Perini | august 2026