Sunday, for Grasp

Your hardest problem isn’t the AI.

It’s trust. An analyst won’t stake their name on a number they can’t trace, so your smartest answer still loses to a spreadsheet.

I’m Carl. I run Sunday, product design for fintech, ex-Swedbank. I couldn’t stop thinking about this one, so I designed the fix. It’s live, just below.

The concept, running

It’s live. Click a blue number, or press ⌘K

How I’d make it trustworthy.

I mapped how an analyst actually works, brief to committee. Grasp wins the first two steps. Trust breaks at the third, where the answer has to survive a room paid to disagree. That’s where I designed.

Get the brief. An MD asks whether the company's growth is real.

Four decisions, and why:

Based on 14 filings · 2h ago High confidence

Growth should decelerate but stay positive. Consensus of ~22% for FY25 looks reachable.

I put the verdict first

The call and how sure it is sit up top, on visible provenance. An analyst trusts it or doubts it in one glance, not after a scroll.

FY24 revenue rose 34% to €218M1

1FY24 annual reportp.42 · income statementRevenue

“Total revenue of €218.4M, up 34.1% from €162.8M in FY23.”

I made every number traceable

Click a figure and land on the exact line in the exact filing. Checking happens in place. No PDFs, no taking the model's word for it.

Watch — enterprise churn. Pressure-test this before you commit: if net retention keeps sliding, FY25 lands at the low end.

I flagged one risk, not twenty

A single thing to pressure-test, not a wall of caveats. A model that hedges everything has told you nothing.

Ask a follow-up about this analysis…
Export to memo Add to deck Copy with citations

I made it leave as a memo

Export with the reasoning and citations intact. What gets questioned in the committee arrives already defensible.

Anyone can design the screen. The real work is knowing which one to design.

A bit about me.

I’m Carl. I run Sunday, a product-design studio for fintech. Before this, Swedbank, one of the Nordics’ largest banks. I work embedded, like part of the team, from first research to the final interface. No handoffs.

I tried to get into the product first. You’re keeping access close for now, so I worked from the outside, on your positioning alone. With your users and data, it gets a lot sharper.

Carl Harrisson

“He champions user-centered design without ever losing sight of how it drives real business outcomes. That balance is rare.”

Joackim Zwahlen — UX Lead, Swedbank

That’s the idea.

I made this because the problem stuck with me. If it’s useful, grab 30 minutes below and I’ll show you where I’d take it next. If not, no hard feelings. I’ll be watching what you build either way.