Since ArkiTask was still early-stage, there wasn’t existing research to lean on. I relied heavily on in-depth conversations with the founder and subject matter experts to understand the core pain points and imagine how users would actually engage with the product.

The founder initially shared a set of rough sketches to communicate his vision, which became a starting point. Through weeks of ideation, feedback loops, and iteration, I expanded on those early concepts, evolving both the design and functionality far beyond the original sketches into a more focused, usable product.

Because the client needed a polished prototype to pitch to investors and early customers, I had to design with clear boundaries in mind.

• Time Constraints: With only two months to design and build, I prioritized addressing the core problems first, leaving “nice-to-have” features for later iterations.

• Technical Constraints: The client’s existing infrastructure limited what we could realistically deliver. By working closely with the founder and engineers, I scoped features that balanced ambition with feasibility, ensuring the prototype was both inspiring and buildable.

With only two months to deliver, we needed an MVP that would feel valuable to early users while compelling enough to attract investors. To stay focused, we prioritized task management as the core flow — the heart of the product’s value.

The MVP included essentials: creating projects, adding tasks, and leveraging AI for task clarity, estimates, and assistance. By focusing in on these features, we delivered a flow that was simple, testable, and showed investors the potential for scale.

Initially, the client wanted every task to require an estimated number of hours before being added. While this would generate useful data for improving accuracy, it quickly became clear the experience was tedious — users had to stop and select hours for every single task.

I advocated for a more balanced approach: tasks and bundles could be added instantly, with estimates generated automatically and editable later. This kept the business need for data intact while making the experience far quicker and more flexible for users.

Other changes and refinements of the Task Selection screens are annotated below: