GroDth

Growt+Depth. This CLI application harnesses the power of CV, Selenium and Jina to help you grow and cultivate your online presence as a content creator.

Comment

Inspiration

There are different components to this application. I got, the overall idea from the power the hashtags have on any given social media platform. Using the right hashtag could be the difference from being Viral to not being seen on the platform. You should always try to use the correct hashtags and that's where this project is aimed to help out. I follow a lot of youtube commentary channels. The ones whose original content is to roast/extend the already available content. That leads to some really obvious copyright issues. However, most of these videos fall under Fair use which Youtube is supposed to protect but in most cases, it fails very badly. So, to help out channels battling the copyright strikes (btw, if you get three of these your whole channel gets deleted and hence this makes it an even bigger threat) the second component of this application was created.

What it does

GroDth is a CLI or Command Line Interface application. At the core of the application, it helps out budding Instagrammers by suggesting them the perfect hashtags to use based on their search or based generally on the use case which they have decided. It also has a whole section for YouTubers. You can have it detect the amount of copyrighted material in your video by giving it the source material. It will then tell you the amount of material you have in common with it. According to the YouTube guidelines. Fair use is usually the case when the amount of copyrighted material is very less. However. there have been cases in which even a fraction of the material could yield to a claim. So far. it only compares the videos and not the audio. I have a plan of adding the audio section as well.

How I built it

I built it using Jina Ai for the search capability. I used Selenium to scrape the hashtags data and I used python to build the CLI. It also uses Cockroach DB for its database. It uses OpenCV to skim through video the For the Instagram section. I went above and beyond in doing my own SEO search. I went through multiple guides on the internet and went through numerous online SEO checker to find which hashtags could be used under which conditions. I have tried to keep the section very general as well as have provided some really good, best practices which you should follow alongside the use of these hashtags for amazing growth. For the youtube section, I went through the various policies that youtube has in place to help protect the rights of the creators. According to the YouTube guidelines. Fair use is usually the case when the amount of copyrighted material is very less. However. there have been cases in which even a fraction of the material could yield a claim. So far. it only compares the videos and not the audio. I have a plan of adding the audio section as well.

Challenges I ran into

There were many challenges that I ran into. To begin with, automation and using selenium is always a pain especially considering after time the whole platform shifted to using react. Using Selenium to automate the scraping turned out to be a bigger deal than I first have imagined. The second and the biggest challenge I faced was the use of Jina AI in the project. Step one, I had to get WSL up and running since I have a windows device. Upon that, the limited amount of resources and guides made it all the more difficult for it to get up and running. I had to look at a number of different implementations. Go through legacy, as well as the recent examples to get it to work. I still wasn't able to make jinabox to work and hence I was only confined to the use of CLI to make it work. For some reason, the post request wasn't working as I wanted it to. I did manage to fix it but it was towards the very end of the hackathon. Not enough time to spin up a good looking frontend to go with it (Jinabox still doesn't work for some reason) ( I'll make it work for the next one)

Accomplishments that I am proud of

The fact that all the various components work and they behave as I wanted them to.

What I learned

I learned a lot. Like really a lot. I had to use so many different technologies in a way that it seemed organic and hence there was a lot of learning. I learned Jina, Cockroach, CLI, Python, OpenCV, automation.

Built With

Share this project:

Updates