Exam Gen-ie

• 

  Initial view, let's create a new multiple-choice exam

• 

  Enter a topic. Questions will be generated from (previously uploaded) documents. You can see options and correct answer

• 

  You can upload PDF documents, page chunking is applied.

• 

  You can export the exam in PDF format

• 

  Source documents that were used for exam questions grounding (citations, references)

Inspiration

The inspiration for Exam Gen-ie came from observing the challenges educators face in creating high-quality exams. The traditional process is time-consuming, repetitive, and often lacks the flexibility to adapt to different learning levels and styles. We wanted to leverage AI to create a tool that could alleviate these burdens, allowing educators to focus more on teaching and less on administrative tasks.

What it does

Exam Gen-ie is an AI-powered application that generates comprehensive, varied, and tailored exams in minutes. By analyzing a document (let's say some chapters of a book or study guide), Exam Gen-ie creates balanced and fair assessments that align with educational goals. You provide a topic and the app provides a set of relevant questions about the given topic. It is helpful for students who want to reinforce their learning and for educators who would like to create tests for students or sample evaluations.

How we built it

We used the following technologies following the RAG architecture:

• Azure OpenAI
• Azure Storage Account
• Azure AI Search

The app is a .NET MAUI solution, so it works on Windows, Android, iOS

Challenges we ran into

Building Exam Gen-ie presented several challenges:

• Ensuring the accuracy and relevance of the generated questions required continuous refinement of the system prompt.
• Balancing the variety and fairness of questions to avoid repetition and bias.
• Designing a simple but user-friendly interface that accommodates the diverse needs of users.

Accomplishments that we're proud of

We are proud of several key accomplishments:

• Successfully developing an AI model that generates high-quality, relevant exam questions.
• Usability enhancement.
• Receiving positive feedback from educators who have significantly reduced their exam creation time.
• Creating a tool that not only saves time but also improves the quality and fairness of assessments.

What we learned

Throughout the development of Exam Gen-ie, we learned:

• The importance of continuous feedback from educators to refine and improve the application.
• The complexities involved in developing AI models that can understand and generate educational content.
• How to balance user interface design with powerful backend functionality to create an effective tool.

What's next for Exam Gen-ie

Looking ahead, we plan to:

• Expand the range of subjects and educational levels that Exam Gen-ie can handle.
• Incorporate more advanced AI features, such as adaptive learning paths and personalized question recommendations.
• Enhance analytics capabilities to provide deeper insights into student performance and learning trends.
• Continuously improve the application based on user feedback and emerging educational needs.

Built With

• .net
• aisearch
• azure
• c#
• openai