Sylla

Inspiration Professors often have to switch between disconnected tools for planning lessons, organizing materials, building slides, and tracking coursework. We wanted to create a single workspace that reduces that overhead and gives educators more time to focus on teaching and supporting students.

What it does Sylla is an AI-powered teaching workspace for professors. It helps users create courses, organize materials, build modules and lessons, track upcoming lessons, import grade data, generate teaching documents, and create presentation decks with AI. It also supports settings, profile management, and downloadable outputs such as PDF and PPTX.

How we built it We built Sylla as a full-stack web application using a React frontend and a Node.js backend. The frontend powers the dashboard, planner, materials, settings, and slide workflows, while the backend handles authentication, course data, persistence, exports, and AI generation. We also integrated Gemini API features for safe academic content generation and added backend file handling to support real uploads and local storage.

Challenges we ran into One of the biggest challenges was turning a polished interface into a fully working platform where every major button and workflow actually did what users expected. We also had to solve formatting issues for generated presentations, especially making exported PDFs readable and keeping preview modals usable. Another challenge was designing AI features that were helpful, structured, and safe for academic use.

Accomplishments that we're proud of We’re proud that Sylla became a real end-to-end product instead of just a prototype. We connected the backend, removed dummy data, enabled real uploads, made planner modules editable and reorderable, tied lessons to upcoming schedules, and added AI-generated teaching documents and slides. We also improved the product’s branding, visual consistency, and export experience.

What we learned We learned that good product design is not just about appearance. A tool feels polished when its flows are complete, reliable, and intuitive. We also learned that AI integration requires careful prompting, safety boundaries, and strong fallback behavior. On the engineering side, we learned how to coordinate frontend state, backend persistence, file uploads, and document generation into one cohesive system.

What's next for Sylla Next, we want to make Sylla even more powerful for educators by expanding analytics, improving generated document quality, adding stronger collaboration features, and supporting more robust storage beyond local persistence. We also want to continue refining the AI workflows so they feel more personalized, more reliable, and more useful in real teaching environments.