DEV Community

Michael Lustig -
Michael Lustig -

Posted on • Updated on

Clipper - Use Machine Learning and Twilio to Turn Screenshots into Video Clips

What I built --- Clipper

Category Submission: Interesting Integrations

Walking Through the Code

Demo Link

Link to Code

GitHub logo technoplato / Clipper

Send us a text and we'll send you a clip of your video

Try It


Docker & Docker Compose

  1. Install Docker
  2. Install Docker Compose

Google Cloud Vision

  1. Follow instructions here to create a Google Cloud Vision Enabled Project and download key.json
  2. Open the env file at the root of the project and make WHERE_IS_MY_GOOGLE_CLOUD_PLATFORM_KEY_JSON_FILE match where you downloaded your file.


  1. Use the promo code DEVHACK20 for $20 when you create your account.
  2. Create a new Twilio Account (DISCLAIMER: this is my referral link)
  3. Navigate to your Twilio Console and copy your ACCOUNT SID and AUTH TOKEN into the same .env file from above.
  4. Buy a Twilio phone number and copy it to your pyclipper/pyclipper.ini.sample file.

Finally, change the name of the pyclipper.ini.sample to pyclipper.ini

Run Application



  • Get google application credentials into environemnt where its needed. I need to figure out docker for this

  • Automate ngrok and twilio webhook configuration

  • Install twilio cli

How I built it (what's the stack? did I run into issues or discover something new along the way?)

I learned a ton and I've got to write more about the experience, but right now, just wanted to submit a post with the Dockerized and runnable version of my project.

The stack is built with Flask, RabbitMQ, Google Cloud Vision, and Docker / Docker Compose (if that counts in the stack?).

I want to say thanks to Twilio and DevTo for putting this on. I had a lot of fun creating this project and will use it and continue to improve on it for my own personal experience and enjoyment after the hackathon. Good luck to all!

Top comments (0)