Thena

Inspiration

As finals approached in Germany, I needed to study efficiently but still wanted to participate in this hackathon and hopefully enjoy a bit of summer too. I was overwhelmed by the volume of slides, notes, mock exams, and recordings I had, and I realized there wasn’t a simple tool to turn all that into focused, actionable study resources. So I built Thena: a smarter study assistant powered by AI that helps students like me get more done with less time and stress.

What it does

Thena takes the raw materials students already have and transforms them into:

• Practice tests
• Flashcards
• Concise study guides The goal is to eliminate manual prep work and provide structured, efficient study tools instantly and intelligently.

How we built it

Thena was built using:

• Bolt for rapid full-stack development and deployment
• Supabase for authentication and file storage
• Netlify for hosting and CI/CD
• OpenAI’s API for generating questions, summaries, and flashcards based on uploaded content

Challenges we ran into

• Output relevance: Making sure the AI-generated questions matched the source material closely took iteration and prompt tuning.
• UI formatting: Structuring and displaying the AI-generated content in a clean, dynamic way was challenging, especially given the variety of possible user inputs and unpredictable API output.
• Time: Building something this practical in just a few days meant making fast decisions about features and scope.

Accomplishments that we're proud of

• Built a working tool that processes real student content into usable study resources
• Successfully handled multi-format input (text, PDF, audio)
• Clean, intuitive UI that could actually help someone prep for finals
• Demo-ready product that already improves personal study sessions

What we learned

• Prompt engineering is everything, especially when generating structured outputs like quizzes or flashcards.
• AI can meaningfully reduce the busywork in studying, but only when grounded in real student workflows.
• Balancing product scope with time constraints is essential to keep momentum and clarity during a hackathon.

What's next for Thena

• Support for lecture recordings: Generate transcripts from recordings and allow for AI materials to be created from those transcripts.
• Increased context: Upgrade the API used by Thena to allow for more tokens and, therefore, more information from user input.
• Larger database: Upgrade the database to account for larger file sizes. Specially useful for recordings.
• Premium features: More personalized content, larger file sizes, voice AI, AI tutor, Q&A, etc. All for premium users using RevenueCat.

Built With

• bolt
• netify
• openai
• supabase