On Voogling
I apologize for beginning this with such a corny portmanteau, but it only felt appropriate given the nature of what the rest of this post is going to discuss.
Voogling — (v.) to search within a video
Do you remember when you first used Google? For those of us that were young enough at that time, our lives have been spent growing up with oceans of information at our fingertips. Imagine a world where you actually had to go to your local library, comb through books using the Dewey Decimal System (whatever that even is) and find your information using the appendix. In all honesty, I had to use Google to find out the name of that system and remember how to spell it.
But one resounding question has been left unasked:
How can I search inside, or for content inside, my media?
In the last decade, accessibility to portable cameras and cellular phones have lead to the creation of dizzying amounts of media circulating the web. People take pictures and videos everyday; billion dollar startups deal with sharing 10 second videos or uploading your life as a series of pictures. So why is that I cannot open my video, and simply say “go to that part when Larry says ‘I love blueberry muffins’”.
Even more so, children rely more on online videos like those published by Khan Academy to find and learn new information, with great positive affect.
Inquiry
This is where our idea started, the concept that we should be able to search within videos, and find content without having to remember where in the video an event occurs. For us, events of interest include:
- Narration — Speech inside of the video
- Text — Text within the video. Education videos generally have text as in the form of presentations
- Images — Anything and everything you see: scenes, animals, people, things.
It so happened that this idea was the one we pursued for the Facebook Global Hackathon 2015. So with the idea in my mind, a tech stack chosen, and tons of enthusiasm / excitement, we flew off to San Francisco to bring our vision to life.
The idea was fairly simple, but to make it manageable, and to have a sexy appeal, we targeted the YouTube platform. This is where a majority of people go to find information through videos, and even listen to political rallies or conferences.
After approximately 6 hours of hacking away at this thing, we had created a stable, reliable service to process and query YouTube videos. We even ‘hacked’ our way into the YouTube player and displayed query results as blue colored bands.
This is when the fun began, and when we tried our first test run of the full Inquiry extension, it fell flat on its face, because we forgot YouTube was a site served through HTTPS, and cannot make HTTP calls. Thus, the next 6 hours were spent attempting to secure our DigitalOcean Droplet with an SSL Cert, so that we could make HTTPS calls. This was our first time even working with SSL, and in the process, our SSL cert went through Facebook Messenger and DropBox. The highlight of our night was when we referred to our SSL cert as “jank-ass” due to the amount of seemingly sketchy sites we visited in order to acquire a legitimate certificate.
I haven’t laughed harder than when a Facebook Backend Engineer who was helping us out accidentally dropped the SSL cert into my DropBox. I think that was probably the most insecure SSL cert ever forged by mankind, and I suppose now Facebook owns our SSL cert by virtue of Messenger. Nonetheless, after 5 hours of trying, we started anew, and this time, it finally worked, with an SSL cert that wasn’t sent through every possible file sharing service.
Then, around 5 a.m., we had finished our MVP, and began our website and promotional video, to round out and polish our product.
Fun fact, we got our video background for the website Googling “bullshit hipster videos.”
After nearly 24 hours of polish, double and triple checking, and practicing our presentation, Team USA with our product Inquiry, ended up taking place 2nd at the Facebook Global Hackathon!
Check out our promotional video:
The product is 100% live, and undergoing changes everyday! We even plan on releasing v2.0 really soon! You can find our product and the download link at:
The current state of the project is as a chrome extension which exposes search on YouTube videos, however you must disable your AdBlocker. We are attempting to work around that issue.
Roadmap
We are going to actively attempt to make this a legitimate product, with the ability to process videos as quickly as possible, and even offer the ability to query for content within a library of videos. We are going to even attempt to enter in multiple business competitions in the Northeast, since we all go to school at RPI (Rensselaer Polytechnic Institute).
Our mission is simple, and that is that we want to bring information in the most accessible format to people. The uses are endless; it would make every video an interactive library, and information more accessible than ever!
