Students want a study tool that’s fast, fun, and feels effortless.
Study Bomb uses AI agents to take what you’re reading and turn it into flashcards, quizzes, podcasts, songs, summaries, and answers in one click — making studying feel less boring.
My research team in the UC Davis Master of Science in Business Analytics program built Study Bomb to be the tool students naturally reach for, because it matches the way they already learn and keeps them coming back.
Product Management
Creating the easiest study tool students will actually use every day
One click, no effort — cut out extra steps that make people quit studying.
Meet students where they are: works on top of the sites and apps they already use.
Feels rewarding instantly: every action gives progress they can see.
Low mental load: keeps focus without overwhelming with features.
The strategy was simple: if it’s faster, more fun, and more reliable than what students use now, they won’t go back.
Cost: We relied on free AI usage, which ruled out paid models from OpenAI or Anthropic and limited how much processing we could do per day. This shaped the scope of features and forced efficiency in product strategy and design.
Time to market: New AI products everyday, and being first matters. Pre-launch, similar apps were released like Microsoft Edge Copilot, Perplexity Comet, and Arc Dia, making speed essential to stay competitive.
Speed: For instant answers, we chose Gemini 2.5 Flash Lite over more accurate but slower and expensive models like Gemini 2.5 Pro. The priority was fast, simple interactions that kept students engaged, even if it meant sacrificing some answer precision.
Limited cloud infrastructure: Most of the app ran directly on-device for speed and maintainability. The cloud was only used for securely storing access tokens and for making requests to Gemini’s API. We didn’t use a cloud database, which meant the app "remembers" past sessions by storing that memory locally.
MINIMUM LOVABLE PRODUCT
We focused on creating a product students would actually love from day one, not just tolerate. Instead of shipping a bare-minimum MVP, every feature in the first release had to feel smooth, rewarding, and worth returning to.
Core focus: The highest-impact features students would love on day one.
One-click learning: Instantly turn any reading material into flashcards, quizzes, or summaries.
Works where students study: Runs on top of existing sites and tools — no switching apps.
Local memory: Remembers user preferences between sessions without sign-in.
What we're launching later to ship faster:
Actions: MCP agents that completes tasks for users
Dashboards: track student progress on a teacher portal
Single-Sign On: Enterprise-grade SSO
Why this worked
Shipping an MLP meant we could launch faster than competing tools, test with real students, and get feedback while the concept was still unique in the market.
Other products best (and worst) features helped make a better app for students
A unique selling point and engaging design is a requirement to attract new users to the shopping app instead of the many competitors users could choose from. The app adds high-retention features while using a design consistent with the most used and familiar apps in the world to make a great user experience.
AI
Shopping
Marketplace
Easy
Checkout
Social
Features
Fast
Shipping
Best AI
Best prices
Local, faster shipping
Most features
No socializing
Many glitches
Difficult checkout
O.K. quality
Cluttered design
Slow loading
