The AI concept that changed our company’s way of working
We changed our design mindset. We transitioned from building all the interfaces and prototypes to allowing others to develop their ideas in a safe place, following our design and product rules.
Fully immersed in AI dynamics
Like almost all technology companies today, we are influenced and affected by AI at our core. Part of this technology adoption has brought new company challenges (ways of work, roles, technical knowledge, etc). Still, it has also led to many positive developments and ideas, allowing us to transform old problems into new opportunities.
Once AI was established as the new technology core, we are playing in a new, unknown field full of opportunities but also uncertainties. How do we build products now? How do we design? How do we work under new rules and roles? Those are questions that we don’t have a solid answer to yet, but in time, we will figure it out. The other side of the story is the opportunities AI will uncover if we choose to adopt this new approach with conviction.
This article explains how we (a Latam HR rising start-up) are innovating from a humble concept, an approach that has been positively adopted and significantly changed the way we work.
From embracing AI to innovating with it
Due to a company event (a hackathon), the UX team and I decided to build something special to present as pre-work. It was a fully functional chat interface made in Cursor for all teams to use and place their ideas, ensuring that whatever they create will maintain all Design System tokens and components. From this simple idea, something bigger started to grow after the event. We wondered what would happen if we decided to build a fully functional dummy platform where everyone could create, play, and share? That’s how the Product Playground was born — a simple tool that has already made us think about how we should work from now on.
What’s a Product Playground?
The concept is straightforward yet robust because of the transformations we are experiencing in this new AI field. The concept of having a place where everyone can “build ideas” connects with the spirit of the building mode feature in video games, where users can construct, customize, and manipulate structures and objects within a game world.
Having a “test or creation environment” in tech companies has, until now, been one that only engineers could use, so enabling everyone to use a test platform to create products easily was disruptive, at least for us.
The definition goes like this (still in progress): The Product Playground is a local environment that “simulates” products in a fundamental way, in terms of their structure and composition (gray boxes and text). The idea is that it can be a place to “play” with quick ideas and functionalities, enabling us to have rapid conversations with teams and users.
Each PM and designer will have their own Playground, enabling them to share with other teams and internal and external clients, thus receiving faster feedback in an environment closely resembling production reality.
While creating this new dummy world with my teammate, David Vega, we realized the potential we could exploit from it, along with the new challenges of protocols and design handoff. To us, those concerns were beautiful problems to tackle if the idea could make an impact, and it did.
How does it work?
https://medium.com/media/74d837ecfd7676b18af2583ebf6cf610/href
Part of the essence of the Product Playground is to be really friendly to everyone outside the engineering team; therefore, we decided to build this Playground with some UX foundations. To start, there is a “read me first” section where users can understand all the knowledge behind this tool, such as the definition, step-by-step setup, templates, design validator, and tool documentation.
Once users understand how to operate the tool (it is effortless, though), they need to pull the repository where the playground lives, and that’s it; you can start to play. We were also cautious in providing some initial prompts to make fundamental changes within the tool, and from there, the imagination is the limit.
Scopes and levels of solutions
We started small, as we always do when building products. So far, we have launched the first version of Product Playground and achieved our primary goal: to spark interest in creating products and sharing ideas among people. We believe this version shouldn’t be the end, so we are planning to evolve the AI tool into something more interesting. Here’s the evolution scope we’re planning:
Phase one:
Mission: “Play and Share”
Contains the following features:
- Local environments
- Local component library in Cursor (Design System tokens + components)
- Playground publishing
- 100% responsive
Phase two:
Mission: “Real components, real environments”
Contains the following features:
- React environments (Collaborator + Admin)
- Actual company library (MCP: Figma + tokens + components)
- Playground publishing
- 100% responsive
Phase three:
Mission: “Company Standard Code”
Contains the following features:
- React environments (Collaborator + Admin)
- UBITS library (MCP: Figma + tokens + components) + rules
- Playground publishing
- 100% responsive
- Code reviewed and optimized for engineering.
We believe that with this AI tool or concept, our way of understanding product building has changed. From our company’s perspective, we see a lot of benefits to highlight, such as:
- Functional designs and ideas that can feel more real in their iterations.
- Speed in idea validation (acceleration stage).
- Shift from the “traditional” mindset that says I must rely on third parties to validate or create something.
- Independence when building product experiences from an idea.
- Engineering inputs; in more advanced phases of the Playground, the components and layouts will be the official ones from the company libraries.
We acknowledge that not everything is rainbows and sunshine. With this new way of building products, we anticipate some big questions and challenges, such as the role of a designer as we know it (I’ll explain it in a few paragraphs) and whether the new handoff method will be a Product playground link. What should happen with Figma files? Do we even need it since now we have a new construction platform? The reality is that we’re discovering answers as soon as we move with this idea. I hope to get you an answer in my next article.
A new design mindset?
As I mentioned at the beginning of this article, we’re experimenting with a possible design mindset shifting process; how do we know? With this new tool, interface design is not just a job for professionals; it’s for everyone. As a result, our design duty may shift to something unfamiliar: the design of systems, or as we’ve started to call it, Infrastructure design. We’re not pretending to add another design role to the already confusing list in the industry, but in scenarios like this, naming what we are building gives us ownership of ideas and processes.
Infrastructure design, as we understand it, involves the planning, design, and maintenance of a large product platform where others can come to build. Our priority right now is to provide all the building blocks (Design system, user experiences, interactions, etc.) to allow others to create. So immediately we launched the Product Playground, a worry started to circle our minds; with this tool, our design job will disappear? This is a valid question due to the significant hysteria in the design community since AI has become a permanent presence. My answer to that is no; there are still many things a designer can do that no one else can. The issue until now is that we have allowed companies to pigeonhole our work as only an interface exercise, and I believe it is more than that.
At the end of the day, if someone is creating stuff in a Playground, it is the designer who decides if such an invention is easy to use, coherent with user needs, and meets other relevant criteria.
Along with this “new concept” that only applies to design teams, there is another one with a broader scale that we are adopting, and I’d like to share: functional design. This concept refers to the Playground outcome, where ideas are essentially functional and ready to implement. We’ve abandoned static design and limited prototypes for rich interactions and complete flows that can be made in minutes.
When we talk about design now, we can see things working in a real environment and have more effective conversations with users and stakeholders. The sense of reality is a plus we haven’t had before.
Another topic we have been navigating and evangelizing within the design team is embracing “designs” from others, especially from those who are not designers. To clarify, the Product Playground is not open to everyone in the company, but only to people in the tech department (PMs, sales, data, engineers, QA, etc). The message is simple on this concern: What was once an insult to us as designers is now real collaboration. We can exchange design and experiences, and also teach them how to do it in the process.
The new company’s way of working.
Since we started embracing the “production acceleration stage” in the company, we see product design with fresh perspectives, especially with the Product Playground in the conversation. It may not be the right or only path to follow in the industry, but for us, it’s been quite a change.
The most compelling evidence we have to validate our hypothesis is time. Before all these concepts (Acceleration stage, Product Playground, Infrastructure design, and functional design), building an idea went like this: Ideation → Model → Prototype → Engineering. The cycle was slow, expensive, and discouraged innovation, as typically only one idea was developed. Now, a process that used to take weeks only takes a few hours. Plus, we can test multiple versions of an interface quickly, preventing ideas from falling by the wayside.
The AI concept that changed our company’s way of working was originally published in UX Collective on Medium, where people are continuing the conversation by highlighting and responding to this story.
This post first appeared on Read More