• 

  Architecture Diagram

AI/ML Track

Inspiration

We came into this competition wanting to win, and this competitive spirit hasn't left us yet. However, as we analyzed the tracks and decided on AI, we based this decision on which track would bring forth the most personal growth and challenge.

What it does

RevNav is an AI chatbot that is the perfect tool for prospective George Washington University students. With this intuitive web application, one can browse dozens of GWU pages with only a few words. From campus life to the academic calendar, RevNav can provide relevant information on all things Revolutionary (GW's moniker).

How we built it

Our team utilized Microsoft Azure resources like OpenAI, Search Service, and App Service to tie together Natural Language Processing and Generative AI. We were also able to connect specific, relevant information to the model from scraped files collected directly from the George Washington University website.

Challenges we ran into

There were many walls that we ran into along the way. Some we were forced to climb, like when the startup code for the Azure resource group wouldn't run, while other walls we had to go around. This would happen when we tried to implement additional resources like Azure Cosmos DB or more complex web scraping techniques with site mapping.

Accomplishments that we're proud of

We are proud of overcoming the rigorous process of troubleshooting and debugging tech that we were seeing for the first time, ever. We gained enough of an understanding of something that was once incredibly opaque to run and build on a useful implementation.

What we learned

We learned a lot about the full-stack development process. This included time management, setting reasonable goals, and gaining the ability to pivot and switch to goals that would be accomplished within a certain time constraint. These are all useful skills for us.

What's next for RevNav

We would like to expand RevNav's capabilities by returning to the goals we had dropped due to time constraints. This includes the more advanced web scraping techniques and implementation of the Azure Cosmos DB. Another interesting resource we explored within Azure was the Azure AI Studio.

Built With

• azure
• css
• html
• powershell
• typescript