Inspiration

Frustrated by the inefficiency of repetitive tasks at the Tier 1 Help Desk, my aspiration in specialized cybersecurity fueled the creation of Project Auto-Remediate, an AI-powered solution designed for autonomous diagnosis and auto-remediation that acts as a crucial force multiplier, redirecting specialized talent toward critical, evolving threats.

What it does

The user inputs any error code, error name, or faulty code and the AI powered website will reply with an explantion of the possible issues and the solutiosn to each one. Additionally, if the specefic error is unknown, there is a separate input box where you can describe any technical issue you are experiencing and the site will provide all the possible diagnosis and the next step you should take for each one. The site also features a chatroom where IT professionals may collaborate with the community for best practices.

How we built it

We used Firebase to store and analyze the real-world help desk data, which allowed us to test and refine the diagnostic process powered by the Gemini API for NLP. Once the Gemini model accurately identified the issue, we created an automation layer to securely trigger validated corrective scripts through remote management tools for immediate bug remediation. We connected the front end and back end through coding in VSCode with HTML.

Challenges we ran into

Our main challenge camefrom our inexperience with Firebase, which complicated real-time state management. We faced difficulties designing the architecture to properly call the Gemini API and manage its asynchronous output. The final challenge was using the Firebase database as the trigger mechanism for launching our corrective automation program. As it was a new concept to us we had to put many hours of research into this project to learn how to properly run our site.

Accomplishments that we're proud of

Our main source of pride in this projection is the fact that we made a working project with tools we have no experience in by extensive research and continuously builidng on each others' ideas.

Built With

+ 2 more
Share this project:

Updates