LearnForgeLite

• 

• 

• 

Inspiration

As university students, we realised how overwhelming it can be to keep up with fast-paced lectures, endless PDFs, and revision notes scattered everywhere. We wanted a tool that not only summarises content but also helps students actively learn, not just passively read. LearnForge was created to make studying faster, smarter, and more personalised, especially for students juggling multiple modules.

What it does

LearnForge is an AI-powered study assistant that helps students revise efficiently by offering:

✅ AI Note Summarisation

Paste notes or upload a PDF, and LearnForge instantly generates a clean, easy-to-understand summary.

✅ Smart Quiz Generation

The app creates personalised Open-ended quizzes based on your notes

✅ PDF-to-Notes Extraction

Upload lecture slides or handouts—LearnForge extracts the text and turns it into revision-ready material.

✅ Downloadable Exports

Students can download:

Summaries (.txt)

Quizzes (.txt)

Everything is clean, simple, and student-friendly.

How we built it

LearnForge was built using:

Backend

ASP.NET Core MVC

C#

OpenAI API (GPT models for summarisation + quiz generation)

PdfPig for PDF text extraction

Frontend

Razor Views

Bootstrap 5

Custom baby-pink UI theme for a calming, student-focused experience

Architecture

Controllers manage PDF uploads, API requests, and exports

Services contain all AI logic and text cleaning

Models define user inputs and AI output structure

The app follows a clear MVC structure, making it modular, scalable, and easy to extend.

Challenges we ran into

We faced several unexpected challenges:

❗ AI Integration Issues

Connecting our local environment to the OpenAI API initially caused endless 401 errors due to:

Environment mismatches

Keys tied to the wrong account

Missing development JSON config

❗ JSON Response Parsing

Large language models sometimes returned text wrapped in invalid formatting (e.g., ```json), which caused deserialization failures. We built a custom cleaning pipeline to strip markup and extract valid JSON.

❗ PDF Extraction

PDFs often contain messy formatting, line breaks, and odd spacing. We implemented normalization steps to improve analysis consistency.

❗ Routing Conflicts

Duplicate Index routes created ambiguous endpoint errors. We refactored methods and removed collisions to stabilise the app.

Despite these challenges, we built a smooth, fully functioning system.

Accomplishments that we're proud of

Built a complete AI-powered learning assistant from scratch

Integrated file uploads, AI, and quiz generation

Designed a polished, aesthetic baby-pink UI to make studying feel welcoming

Implemented flexible quiz controls (difficulty, question type, question count)

Achieved smooth PDF → AI → summary/quiz workflow

Overcame OpenAI key + configuration issues that blocked development

The final product is clean, fast, and genuinely useful for students.

What we learned

Throughout this build, we learned:

How to design and structure an ASP.NET MVC application

How to interact with external APIs securely

How to parse & validate JSON returned by LLMs

How to implement file uploads and PDF text extraction

How to solve API authentication, routing, and serialisation issues

How to build user-friendly, minimalist UI inside Razor views

Most importantly, we learned how to turn a rough concept into a polished educational tool.

What's next for LearnForge

We already have a roadmap for future improvements:

🎯 MCQ Auto-Marking

Students answer quizzes → system grades them → provides explanations.

🎯 Study Progress Tracker

XP system, streaks, level-ups, and performance analytics.

🎯 Flashcard Export

Automatically generate Anki-ready sets.

🎯 Module-Based Organisation

Organise summaries/quizzes by course or topic.

🎯 Study Timer + Focus Mode

Pomodoro + distraction-free UI.

🎯 Interactive Quiz Mode

Click-to-reveal answers + instant feedback.

LearnForge has the potential to grow into a full AI-driven learning platform built for students, by students.

Built With

• asp.net
• bootstrap
• c#
• css3
• googlefonts
• html5
• javascript
• openai
• pdfpig
• system.text.json
• visual-studio