autoM8 | Devpost

Inspiration

Our inspiration for developing this tool stemmed from the challenges we faced with handling large files. Whether it was struggling to process extensive datasets for model training, grappling with unwieldy CSVs during data cleaning without a concise summary, or uploading files to the cloud for basic image processing—these tasks were time-consuming and often overlooked. Editing these files manually was both labor-intensive and inefficient. This is why we decided to enhance the already impressive MASV tool with a new extension designed to streamline these processes.

What it does

With the help of MASV, users are able to apply edits to large amounts of image and video files, while guaranteeing the speed, efficiency, and consistency that MASV provides. While focused on images and videos in this setting, the concepts are easily extended to virtually any file processing text you can imagine.

How we built it

The web app was built and deployed with Defang and the file-saving system was created using MASV API. The image and video processing use python OpenCV library and cloud integration is done on Google Drive.

Challenges we ran into

One of our biggest challenges was switching projects with less than 24 hours left during the hackathon. With this, it was difficult for us to catch up to the level that other teams were ahead in their projects. For this reason, we were unable to finish the project to the point that we envisioned. It lacks the fully automated features it could have with more work, but the components are there for further polishing.

Accomplishments that we're proud of

Still managing to get a product together despite the rough start. Managed to stay persistent and not lose our desire to build something cool. Ultimately we are really proud of the idea and implementation we have put together and see it as something we can build our own workflows on top of.

What we learned

Learned how important it can be to remain open to new ideas and to pivot quickly. If we hadn't made the quick decision to create a different project, who knows what kind of product we would have now. We're taking away this idea of flexibility and looking to apply it creatively in our future projects.

What's next for autoM8

Completely automate the data processing pipeline. Provide more support for file formats and processing tasks, making it more applicable across domains. Integrate unique processing featues (LLM analysis, more sophisticated classification of images, etc.)