25: Learning Machine Learning with Sophia Ciocca

"Trust that you will figure it out"

Another guest! Should I just make this all guest episodes??

I first heard of Sophia from my Hacker Noon news feed when her article titled "Spotify’s Discover Weekly: How machine learning finds your new music" went viral. Imagine my surprise when I found out she was a Grace Hopper fellow and in the same building as me for the past three months! This is the first Fellow interview we've had, and we go deep on both Machine Learning and the Impostor Syndrome that comes with learning, and then teaching. Stick around to the end for a special bonus on where Sophia ended up!

Show Topics

  • Sophia's Origin Story from Economics/PreMed to Teaching Fellow
  • The origins of Sophia's viral blogpost
  • Sophia's Writing Tips
  • Researching Spotify from outside Spotify
  • Public perception around Machine Learning
  • The Three Recommendation Models Spotify uses
    • Collaborative Filtering: "Just" matrix math!
    • NLP: Song Metadata + Crawling the web => vectors
    • Audio: Predicting if you'll like totally new songs
  • Research as a Job-Search Strategy?
  • Pandora's Human powered recommendation vs Spotify's Machine powered learning
  • Sophia's history of teaching coding
  • Deciding to learn code despite gender barriers and a late start
  • The benefits of being a Teaching Fellow
  • Tips for teaching coding?
  • What Sophia is Working on: Personal Portfolio! Trying out GraphQL
  • Stick to the end for the postscript on where Sophia Ended Up!


Subscribe on Apple Podcasts or any RSS based Podcatcher, or download episodes at impostor-syndrome.org! Send your questions and feedback to @swyx or swyx@swyx.io. If this story reminds you of someone you know, send it to them!