Inspiration
The main insipiration for this project was definitely BracketHacks 1.0, and also Symbl.ai, their services all gave a good outlook on ML applications.
What it does
ITransCrypt is a video transcription platform. It takes a video file from the user as input and then gives them a transcription for that file, also provides them options to download it directly as a .txt or a .json file.
How I built it
There were several stages to building this application: 1) Working on the backend to store the video file in the backend + Create a URL instance in the MongoDB Cloud database. 2) Creating a frontend using ReactJs 3) Creating Symbl.AI API calls for transcription 4) UI and downloading features for the transcription.
Challenges I ran into
There were a lot of issues that I faced because I have never worked with video files as form data. I had to use various new modules like multer for storing videos etc. So, initially it was hard but yeah I solved the errors. Next were two main issues that I faced one being fetching URL from MongoDB and making API calls and getting transcriptions and of course during deployment on Heroku( a lot of errors!).
Accomplishments that I'm proud of
This was a new type of project for me so I am very satisfied with the site I created, especially the audio bot, even though it was easy to implement it was a completely new thing to me, so yeah I am proud of it.
What I learned
There were a lot of learnings:
- How to create backend for storing videos
- How to store URL of those videos on MongoDB
- Working with video file in Form-data
- Symbl.AI APIs
- Heroku deployment for MERN apps ## What's next for ITransCrypt-Web-App Even though I have a functioning website which allows the user to upload their video files but since I don't have the database paid function I can't store those video files, so for now I am only using a dummy video URL to generate transcription. I will work on improving this in the future.


Log in or sign up for Devpost to join the conversation.