Building a to-do app isn’t the hard part. Building one you actually want to use every day—that’s the challenge. The InfiniteUp to-do app team (Barrett, Wasim, Malaz, and Hiba) is tackling it by doing less on purpose: fewer features, faster feedback, and a ruthless focus on the core loop.
“We should focus more on InfiniteUp,” Barrett said early in the call. The plan: ship the basics, polish the touch points, and only then think about extras.
Ship the core first: the heart of the InfiniteUp to-do app
The team keeps coming back to one truth: task creation has to feel effortless.
“I think the most effective thing to do at the moment is creating the tasks,” Malaz said. Barrett agreed: there are two paths—fast add on the home screen and a full ‘more options’ flow. The moment you enter a task name, background AI assistance kicks in to propose subtasks.
Malaz outlined the AI pipeline: a quick check—does this task need subtasks?—and if yes, generate titles plus start date, due date, priority, and category for each item. “There should also be a start date for the task,” Barrett added, so the daily view can prioritize intelligently.
They’ll use a fast model for snappy responses. “We can use the fastest one because it’s really like a yes or no,” Barrett noted.
Learn how InfiniteUp approaches AI building in Why InfiniteUp Focuses on Gen AI APIs — Not Reinventing the Wheel.
Minimal copy, consistent design patterns
Language is kept intentionally short. “I would just have ‘add a task,’” Barrett said. “Least text possible.”
Consistency gets the same treatment. Wasim defined simple animation rules they’ll reuse everywhere:
- App bar elements slide down with a subtle fade.
- Bottom calls-to-action slide up with a little extra emphasis.
- Page content cascades up with short, staggered fades.
“I like where it’s a bit delayed for the second piece of the animation. Reminds me of Apple,” Wasim said. Haptics follow the same logic. “Let’s do medium and see how that goes.”
The Today view: time-boxed or just priority?
The team debated two versions of the day view: an hour-by-hour schedule vs. a single prioritized list you can reorder.
Hiba likes seeing time: “Time is important to see… I prefer to see the time.” Barrett’s take is simpler: hide the clock if you want and let AI order the day. They’ll likely ship both behind a setting. “It’s not about versions,” Barrett said, “it’s more like the ordering of what we do.”
Components, not pages: InfiniteUp’s seamless flow
To make transitions feel seamless, they’re leaning on shared components. Tap More options on a quick add and the same title field appears on the next screen—no visual jump.
“Watch how they do it,” Barrett said while reviewing a reference. “If you have the same component on both pages, that’s the only core thing you actually need.” Conditional visibility handles the rest (the big plus hides once you go deeper).
Personal touches, not mascots
Rather than invent a character voice, they’ll keep things human. A tiny toast after fast add, addressed to the user: “Barrett, your task has been created.” As Barrett put it, “Make it more personal,” but don’t add friction.
Quiet polish, fast iteration
Hiba and Wasim tuned animations on login, onboarding, and signup—keeping it light. “We don’t necessarily want to animate every single piece,” Wasim said. “Be subtle while visible, while elegant.”
They’re also wiring a View More → All Tasks page with simple filters (due today, past due, future, done; category; priority). Nothing fancy—just enough to find what matters.
A small team moving like a big one
Roles are clear and handoffs are thin. Wasim shapes and stitches the UI, Malaz implements the AI logic and data, Hiba sanity-checks the experience, and Barrett trims scope, pushes for clarity, and locks language.
“It’s not the end of the world if we can’t do it,” Barrett said of ambitious ideas, “but I feel like we’ll find a way.” The guardrails are simple: fewer features, tighter feedback, and polish where it counts.
If you want a proof that four people can build something sharp without drowning in features, keep an eye on InfiniteUp. The team’s bias is clear: get the core right, and let everything else wait.