Snippit

GIF
Homepage into loading page
Your video is ready!

Inspiration☁️:

Our inspiration for "Snippit"✂️ comes from the infamous youtube meme, titled "Barack Obama Sings Call Me Maybe!" The basis of our app focuses on producing a "meme" according to the users request and promotes our enthusiasm for learning new concepts like Text-to-Speech recognition and the overall journey in completing a hackathon. We agreed that in a competitive environment like Hack the North, it is equally valuable to bring fun and laughter to all hackers at such a memorable event!

What it does💁‍♂️:

"Snippit" is a real-time application that allows the user to translate any sentence of their liking into a randomized video file, saying exactly what the user requested. It uses a combination of API's as well as a tailored database that consists of more than 100,000 words with matching timestamps and YouTube URL's to ultimately stitch together a short but funny clip!

How we built it🔨:

Frontend: React, CSS
Backend: Golang, SQL
Services: CockroachDB, AssemblyAPI

Challenges we ran into 😳:

Devising an unbiased and efficient algorithm to populate our SQL database with words from a transcribed text.
Understanding FFMPEG source code to adapt with Golang to interact with multimedia files and modify them accordingly.

Accomplishments that we're proud of💪:

Successfully coupled multiple API's with backend gateway including AssemblyAPI.
Working and communicating effectively with the team, engaging in all aspects of engineering the application including UI/UX Design, Database Management, Full Stack Development, and API Integration.
Experimented with multiple API's to reduce download time for YouTube videos.

What we learned🧠:

Working with new languages and tools to build both the frontend and backend of the application.
Learned how data processing works and how to engineer data for feeding it to a language processing service.