The dashboard of Bloomscroll
Uploading content for a course to generate reels for
The course progress list of topics
A theory reel, designed for teaching content
A test reel, where the user got it right, reflected in the increase in the progress bar
A test reel, where the user got it wrong
Bloomscrolling
Bloomscrollers

🌱 Bloomscroll: Studying While Doomscrolling

Inspiration

Bloomscroll started with a very honest realization:
We all wanted to doomscroll but we also needed to study.

Scrolling is effortless, familiar, and strangely comforting. Studying, on the other hand, often feels rigid and demanding. Instead of fighting that instinct, we asked a different question:

What if scrolling itself could become the study method?

Rather than resisting the doomscroll, Bloomscroll embraces it, adapting the habitual vertical scroll into a structured learning flow. Each scroll pushes the student forward cognitively, not just visually. Our goal wasn’t to eliminate distraction, but to redirect it to something positive.

What We Learned

Building Bloomscroll taught us that learning tools succeed or fail based on alignment — between user behavior, technology, and design.

We learned:

Habits are more powerful than motivation
Structure matters more than volume
Learning systems must adapt to how people already behave, not how we wish they would behave

We also learned the importance of collaboration under pressure: ideas improve rapidly when they are tested, critiqued, and rebuilt as a group.

How We Built the Project

Bloomscroll was built using a combination of:

Gumloop AI to process and generate learning materials
MongoDB Atlas to store user progress, states, and interactions
Svelte and SvelteKit for UI structure, routing, and integration with the backend
TailwindCSS for the app's look and feel

Rather than designing everything upfront, we worked in fast cycles:

build → test → critique → rebuild
logic → UI → logic again

This allowed us to adapt quickly as we better understood both Gumloop’s capabilities and the constraints of a hackathon environment.

Challenges We Faced

Our biggest challenge was understanding and reasoning through Gumloop’s logic model. Translating our conceptual learning flow into a working AI-driven pipeline required experimentation, failed attempts, and rapid rethinking.

We also faced the inevitable time pressure of a hackathon:

limited time to debug
limited room for feature expansion
constant trade-offs between polish and functionality

Despite this, we persevered. As a team, we leaned into collaboration sharing responsibilities, communicating clearly, and supporting each other when things were tough. What could have been blockers became moments of collective problem-solving that Bloomscroll survives to tell the story of.

Reflection

Bloomscroll is the result of persistence, creativity, and a shared vision.

What began as a simple desire to study whilst scrolling evolved into a working product that reimagines how learning can fit naturally into contemporary digital habits.

More than the final app, the project represents what’s possible when a group commits to adapting, learning fast, and building together under pressure.