The Challenge

A startup reached out with a tight deadline and a big ask. They needed a complete front-end for their new API management platform. They had less than 30 days to build it. And they had a major conference coming up where they would show the platform to potential clients and partners.

The stakes were high. The platform helps companies manage real-time data streams and transform raw operational data into business-ready products. Under the hood, it connects request-based APIs with high-throughput, real-time systems. Designing a front-end that makes all of that feel coherent and usable is no small task.

The timeline made it even harder. Most projects like this take months. They had weeks.

The Approach

We started with what we already had: our design system. The atoms and molecules were ready. The coded components existed. We did not need to build buttons or form fields from scratch. This freed us to focus on what mattered most, pages and user flows.

AI made the difference. Designers used AI to understand complex concepts fast. The platform deals with streaming protocols, certificate management, and data governance. These are not simple topics. AI helped designers grasp how the system should work and figure out the right flows and patterns.

Developers used Figma MCP and AI coding tools to bring designs to life. We have AI agents that produce high-quality front-end code from Figma designs. Our developers guided the tools, reviewed the output, and refined the results. Designers and developers stayed in sync.

We focused on simple flows first. The platform has many features, so we picked the core paths users would take. We built those first, shared them early, and adjusted fast based on feedback.

The backend work was done by the startup's own team. Our front-end connected to their APIs. This split let both teams move in parallel.

What We Built

The platform lets users write API specifications, manage products, and control access in one place. We built a complex editor where users create OpenAPI specs with support for both JSON and AVRO schemas. The interface guides users through spec writing without hiding the technical details they need.

Once specs are ready, users manage and deploy them as API products. The catalog gives teams a clear picture of what APIs exist and what state they are in.

We designed a hierarchical structure for applications and subscriptions. Subscriptions control how consumers access the APIs. This lets teams roll up metrics and manage authentication in flexible ways. We shipped API key authentication for the first release. mTLS certificate management came after the conference.

We created role-based access control. Users, groups, and roles have different permissions at multiple levels. Admins can invite team members, set up groups, and manage who can do what. Multi-environment support lets organizations keep dev, test, and prod separate.

We translated complex security and enterprise features into clear choices. The interface keeps hard tasks simple.

The Result

We delivered the core platform in time for the conference. The startup showed a working product. The response was strong. Several interested clients emerged and new partnerships formed.

After the initial sprint, work continued. We refined features and added new capabilities. The foundation we built in those first weeks supported everything that came after.

What Made It Work

The design system saved weeks of work. AI tools made each person more productive. Early communication prevented waste. Sharing work before it was perfect kept the team from building the wrong thing.

Disclaimer

Due to confidentiality agreements, we cannot disclose the client's name or product details.

Read another case study:

Creating Quave One: a product designer's perspective on creating a Cloud hosting platform.

Read Case Study →