📚 Syllabus AI
🚀 Inspiration
Our journey began with a universal student challenge: the syllabus. For many, it's a daunting document—a wall of text with vital information buried in legalistic jargon. We've all experienced the frantic search for a due date, the missed office hours, or confusion over grading policies.
We realized this manual, time-consuming process is a source of academic stress and a barrier to proactive planning. This led to our core inspiration:
What if we could democratize this information—making every key detail of a course instantly accessible and understandable?
We saw an opportunity to transform the static syllabus into an interactive, dynamic tool, directly addressing the UN’s SDG 4.4: increasing youth skills for employment, decent jobs, and entrepreneurship.
💡 What It Does
Syllabus AI is a multi-feature platform that transforms any syllabus into a personalized study hub. Here’s what it offers:
🎯 Intelligent Summarization
Extracts and highlights key details like instructor info, course objectives, and grading policies.📆 Dynamic Deadlines & Alerts
Auto-generates a customizable calendar with due dates and exams. Push notifications help users stay on track.🧠 Personalized Study Plan
Suggests weekly plans based on assignment weights and course structure.💬 Interactive Q&A Chat
Students can ask syllabus-specific questions like “What’s the late work policy?” and get instant, accurate AI answers.
🏗️ How We Built It
🔹 Frontend
- Built using React for rapid UI development and seamless file uploads + AI interaction.
🔹 Backend
- Django powers backend logic and APIs.
- PostgreSQL ensures secure and reliable data storage.
🔹 AI & ML Pipeline
- PDF Parsing: Extracts raw syllabus content.
- Information Extraction: A fine-tuned LLM identifies course details like dates, names, policies.
- Q&A System:
- Uses a vector database + RAG (Retrieval-Augmented Generation) to answer questions with syllabus-specific context and zero hallucinations.
🔹 Deployment
- Hosted on a cloud platform for global accessibility and scalability.
🧗 Challenges We Ran Into
Unstructured Data Variability
No two syllabi are the same. Regex and basic NLP failed. We solved it using context-aware LLMs + prompt engineering.Accuracy & Reliability
A wrong due date is worse than none. We implemented multi-stage validation with a second model for cross-verification.Novel UX Design
Iterated based on real user feedback. Introduced features like Q&A and calendar after testing the initial MVP.
🏆 Accomplishments We’re Proud Of
- Built a Scalable AI Pipeline that can handle unpredictable syllabus formats.
- Created a Unique User Experience that moves beyond summarization into actionable planning.
- Solved a Real-World Problem that every student faces, with a meaningful tech solution.
🎓 What We Learned
Prompt Engineering is Powerful
Crafting the right prompts unlocked the true potential of our LLM.Iterative Development is Key
Listening to users helped us evolve from a tool to a true assistant.Robustness is Essential
With academic deadlines on the line, our backend had to be resilient and error-proof.
🔮 What’s Next for Syllabus AI
LMS Integration
Plug into systems like Canvas or Blackboard for automatic syllabus ingestion.Document Expansion
Extend summarization to research papers, legal documents, and more.Academic Assistant Vision
A complete student companion for:- Time blocking
- Project planning
- Study group matching
- Time blocking
Our mission: Become the indispensable digital partner for every student.
✅ Built with: Django · React· PostgreSQL · LLMs · Vector Databases · RAG
🌐 Built during a hackathon. Changing how students interact with their syllabus—one upload at a time.
Log in or sign up for Devpost to join the conversation.