mixa-ai

mixa-ai

Isaac Johnson

| Capstone Project

Welcome to the informational hub for Mixa-AI,

A groundbreaking capstone project that reimagines video editing through the power of artificial intelligence. At its core, Mixa-AI is designed to streamline the video editing process, making it more intuitive, efficient, and accessible for users of all skill levels.

About Me

Hi! I am Isaac Johnson a senior student at Neumont College of Computer Science studying Software Engineering.

I have been fascinated with the implementation of AI in software and am currently looking for full time work!

Feel free to reach out to me on LinkedIn or GitHub to learn more about me and my work!

About the Project

mixa-ai is a cloud based, AI powered, video editing website.

Through the use of vision AI, embeddings, and langue models it can understands the content of your videos.

The goals was to create a cloud-based video editing platform where the user can simply describe what they want though text and the AI will do the rest.

Mixa-AI then uses this input and the analyzed video footage, identifying relevant segments based on both the visual content and any accompanying transcript returning a fully edited video.

Project Goals

Overall, the goal of this project was to learn!

I learned a lot during the project but here are the highlights:

  • Create an efficient serverless event driven backend architecture that focuses on keeping costs and response times to a minimum
  • Use and combine new cutting edge AI technologies to create innovative features
  • Create a functional frontend implement real-time features
  • Managing a project from conception to deployment

    • Project Features

  • Upload video content to the platform
  • Mixa-AI then analyzes the video's content and transcript
  • Through a simple text-based interface, users describe the edits or segments they're interested in. For example, "compile all segments where a dog is visible" or "create a promotional video for my product with the following features: ..."
  • The AI identifies and extracts the relevant segments based on the user's request and compiles them into a new, edited version of the video or videos.

    • Is It Live?

    As much as I wish to make this available to everyone, the current decision is against hosting it.

    The cost of the AI processing and the video storage is too high for me as a student to maintain at this moment.

    Additionally, the website for this project, while completely function, was not my top priority during the time given from my capstone and needs improvement.

    Technical Details

    For more technical details please see the README in the GitHub project linked above!