2025-03 · Lead Product Designer

UX Design Foundation

Translating Funda's human-led deck building course into an app experience — the launch-period UX design, starting from AS IS / SHOULD BE deconstruction.

The moment we decided to translate this into an app, the first thing I did wasn’t draw screens — it was deconstruct everything humans had been doing.

Slide Academy started as a successor to “Material Creation Course,” a human-led short-form workshop Funda had been running for several years on Discord, Notion, and Zoom. Real user pain points had piled up along the way. The decision came down in March 2025: “Let’s make this into an app you can learn from anytime, anywhere.” Running in parallel with another project, I had about a month to ship the design foundation.

Under that timeline, the first thing to settle was what not to build.

Design principles

Deconstruct the experience before you screen it

The first move wasn’t drawing screens — it was breaking down the human-run experience Funda had been operating. How people asked questions on Discord, how they read material in Notion, the moments that lit up in a Zoom session — the course experience is a stack of thin, layered experiences. Before mapping any of that onto screen structure, I had to fully take it apart.

From there I laid AS IS (what humans had been doing) and SHOULD BE (what the app should reproduce) side by side. We decided we wouldn’t “port the experience to screens” but “translate it into the app’s native form.” The real-time energy of a Zoom session, for instance, became something different inside the app: a live notification + immediate archive + comments.

Don’t draw until the structure is visible

The order was: sitemap → task flows → user flows → personas → wireframes. I put wireframes last on purpose. Starting from drawings drags the structure toward whatever the drawing happens to look like.

The sitemap settled into six tabs: Home / Study Sessions / Learning Content / Training / Submission Tasks / Q&A. Task flows decomposed every interaction in the course experience from the user’s side — sign-up, joining orientation, viewing learning content, taking a training challenge, submitting a task, posting a Q&A. User flows went down to each task’s branches, error handling, and exit conditions.

I placed three personas: a mid-career sales lead looking for well-rounded skill growth (A), a PM at an IT company who wanted to learn information design (B), and a layer of folks expanding into deck-making skills for freelance or side work (C). Every screen got checked against “who is this for?” with these three as the foundation.

Only after those four layers were locked did I open a fresh file and draw wireframes.

Decide what you’re not building, first

Under a one-month constraint, what to cut had to be the first decision. The cut: building a brand worldview and developing a custom visual language. With no evidence yet that the app experience even held together, pouring resources into branding was the wrong sequence.

Instead, we reused Funda’s existing design language as-is. With a deliberately worldview-less, simple UI as the assumption, the validation scope shrank to one question: “does the experience hold together?” Layering a worldview on later turns out to be faster than building everything at once unverified.

How it actually went

Running alongside another project for that month, I worked the schedule backward from the end date and laid out: AS IS / SHOULD BE → IA → task flows → user flows → personas → wireframes. I showed the PdM and engineer the in-progress output at each layer’s lock-in point to minimize rework.

Tooling stayed deliberately simple. Figma held the IA, flows, and wireframes in a single file; personas got their own. No exotic tool stack. That, too, was part of “decide what you’re not building first.”

What came next

The launch-period “no worldview” call paid off exactly as planned six months later, in the rebrand. Once the app experience had proven itself, in September 2025 we pivoted concept to “the Duolingo of deck building” and added the worldview back on top. Had we tried to build a worldview from day one, the pivot would have arrived much later.

Deciding not to do something early makes it easier to commit hard to doing something later.