Aesop.AI

Inspiration

We came up with this idea while watching the sponsors of KnightHacks speak on stage actually. One thing we noticed was that, even though these were the people offering the challenges and basically controlling the events of that weekend, not many people paid attention to them, only ever reacting to what was on the slides, and even talking over them at times. Because of this we thought, if even employees at the biggest and brightest companies struggle to command a room, what about the people who can't even afford a certification they need for their job to get a raise? Because of this, we decided to create Aesop.AI, an AI public speaking coach that guides people in their journey in becoming great communicators.

What it does

Aesop.AI takes just two inputs from the user. A video of them giving a speech, and a short, two sentence context about what their speech is about. Then, AI Agents (That we like to call, "Data Agents," will split up the video into three different components, A Transcription. An Audio Recording. and a Video Recording. Each of these are individually scored by Data Agents and the user is given values based on their performance in that section, for example, they wrote a very good speech, so their written score is an 8.

Then, another AI Agent will appear, this one is called the "Action Agent" and their job is to take the scores the Data Agents gave the User's and explain to the User what those score's mean, and how to improve within the category.

How we built it

We used a pretty simple tech stack, our Frontend was just HTML, CSS, and JavaScript, and our Backend was mainly coded in Python.

Challenges we ran into

There were a lot of challenges we ran into. For half of us, it was our first Hackathon, and for most of us, we had zero idea what we were even doing in VSCode, let alone when placed in front of an AI agent. A lot of the struggles we faced came from making silly mistakes that most people wouldn't have made with experience, but it was through that experience that we grew as programmers and as individuals. A major challenge that we faced was actually getting our Agents to work harmoniously, we probably changed our system architecture like 4 times before we decided on a system that worked.

Accomplishments that we're proud of

We are very proud of completing a project that not only looks good, but performs good as well. And we are incredibly proud of being able to complete so many challenges along the way as well.

What we learned

We learned a lot about what it actually means to be developers in a crunchtime like environment, 48 hours of straight coding with almost zero breaks was one of the most fun, if tiring things we've ever done. We also learned how to effectively use AI agents to make ourselves more effective in the workforce, and that's a skill that we can take anywhere we go, regardless of whether or not we want to stay in the CS field or not.