Viral.AI

From New to Improved: the Viral.Ai flowchart process of a video

Inspiration

When we first got together as teammates on Friday, we decided to bond by making a Tiktok. This got us thinking about how generative AI could be used in the content creation space.

What it does

Analyze Trends

Viral.AI helps creators understand what kind of video content is currently trending and performing well in their niche.

Optimizing Titles and Descriptions

Viral.AI can suggest relevant hashtags and craft compelling titles and descriptions based on successful data that will grab viewers' attention and increase click-through rates.

Providing Improvement Suggestions

Viral.Ai offers specific recommendations on how to improve the user's video, including editing techniques, content elements, and audience engagement strategies.

Leveling the Playing Field

By providing these tools and insights, Viral.Ai empowers new creators to compete with more established channels that might have larger teams and resources.

How we built it

Viral.AI is built completely in Python. The frontend was constructed using the Reflex library, while the backend was mostly supported by the Google Gemini API.

The Process

Flow Chart

Challenges we ran into

We ran into a lot of challenges along the way, specifically with conceptual theory. For instance, we wanted to employ a regression algorithm with the vector embeddings to predict the number of views and likes a new video would achieve, and calculating an overall "Virality Score". However, it was out of the scope of the two day project so we decided to focus on other features instead. This would be the next step that we want to expand to, as it would mostly involve extracting more data on more Youtube videos and deciding how to process them. Through this project, we realized the potential of vector embeddings in today's world with AI, as a very cheap way to represent large amounts of data to be analyzed and processed.

Accomplishments that we're proud of

For many of us, it was our first exposure to machine learning and utilizing APIs of Gen AI. We are proud that we took the initiative to learn new things, but most of all, we are proud that we made an effective product that can help people all over the world.