Inspiration
In today’s AI-driven world, students have unprecedented access to powerful tools—but many, like high-schoolers, aren’t permitted to use ChatGPT directly. After speaking with a student who couldn’t access raw AI yet flourished when using guided helpers, we saw the need for a friendly study companion. StudyWhale was born to bridge that gap, packaging AI’s capabilities into an intuitive platform that simplifies studying for everyone.
What it does
StudyWhale lets students upload lecture notes, PDFs, or plain-text notes, YouTube links, and instantly provides:
- Concise AI summaries that surface the key ideas.
- Custom flashcards with questions, answers, and memory aids.
- Real-time usage feedback so learners can track their study progress.
How we built it
- Built with Bolt’s no-code tools so we could set up subscriptions and webhooks without heavy backend coding
- Frontend: React with TypeScript for a clean, responsive experience.
- Backend & Database: Supabase for auth, storage, and our extra_packs RPC in Postgres.
- AI Engine: OpenAI’s GPT-3.5 and GPT-4o powering the content generation.
- Payments: Stripe webhooks for subscriptions and on-demand pack purchases.
Challenges we ran into
- Balancing cost vs. quality: Finding the right mix of AI models to keep response speed high without blowing API budgets.
- Reliable way of executing complex subscription plans with a pay-per-use product added to the mix
- Global pricing UX: Displaying USD-based pricing in a way that feels clear and fair to international students.
Accomplishments that we’re proud of
- Working prototype in under 30 days: From zero code to a working prototype that handles uploads, AI processing, and payments.
- Seamless user journey: Built and tested the full cycle—purchase, webhook handling, database update, front-end refresh.
- Modular, no-lock-in stack: A serverless architecture that can be extended to new features or markets without major rewrites.
What we learned
- Got up to speed with Bolt’s no-code environment faster than we expected—it quickly became the backbone of our workflow.
- Connecting Stripe, Supabase and OpenAI showed us how to stitch different services together, from secure sign-ups to seamless payments and AI calls.
- Seeing the difference a clear interface makes—like highlighting “remaining packs” in big text—cut confusion and support tickets dramatically.
- Building our login and session flow with Supabase underlined why solid authentication and state management are non-negotiable.
- Designing the extra_packs table and custom RPC function taught us that a thoughtful database schema pays off big later on.
- Debugging webhook issues across Bolt, Vercel and Supabase logs drove home the importance of thorough logging and error handling when you’re juggling live payments and AI.
What’s next for StudyWhale
- Market pilots: Roll out in US, Canada, and Australia to validate messaging and pricing.
- Multi-currency display: Integrate Stripe’s local-currency features for smoother global checkout.
- Create new student resources like a quiz/test mode
- Create a handy study timer complete with deep-focus “whale sounds.”
- Explore embedding StudyWhale into LMS platforms
Built With
- bolt
- github
- javascript
- lambda
- netlify
- oauth
- openai
- postgresql
- react
- stripe
- supabase
- tailwind
- vercel
- whisper
Log in or sign up for Devpost to join the conversation.