Join usDocs
Education / University of Michigan

Streaming Deepnote to Twitch by a University of Michigan professor

Professor Christopher Brooks had a goal to create an online learning environment for students of his class. So, in addition to teaching at the university, he set up weekly online meetups which students can join either via Zoom for an interactive experience, or on Twitch for rather passive watching. Chris screen-shares his Deepnote project, and over the sessions, the group learns to predict the next Hart Trophy winner in NHL.

On a traditional campus, students could learn at hackathons, meetups, and easily look at their peers' laptops to help each other. However, hands-on learning became challenging when moving online. Standard tasks such as installing libraries, configuring Docker, setting up the environment got more complex. Transition to Deepnote was, therefore, very natural.

Use cases

predicting Hart Trophy winner in NHL, collaboration between students and the professor, extracurricular activity

Data stack

Python, pandas, matplotlib


Christopher Brooks is an Assistant Professor at the University of Michigan and is applied Computer Scientist who builds and studies the effects of educational technologies in higher education and informal learning environments. Dr. Brooks has a particular domain focus on data science education and methodological interests in predictive modeling, learning analytics, and collaborative learning. [1]

University of Michigan about 1
University of Michigan about 2


An important aspect for a student is to be able to reproduce any result reached during the session with the professor. Deepnote allowed the group just that: with a one-click Clone button, students could, on their own, replicate and experiment with any results obtained previously.


Dataset creation

Deepnote allows the group to build a dataset shared across all projects in the team. This way, even all individual projects of students could access the commonly curated most up-to-date data.


Collaboration better than in person

In an online world without Deepnote, Chris would screen-share his locally-running Jupyter instance. But as students started asking deeper questions, sending snippets of code over the chat was inevitable and remote help was almost impossible.

Deepnote not only solved all this but went even further – multiple students could edit a notebook at a time or seamlessly alternate based on who was leading the conversation, which proved an incredible help.

Deepnote created a sense of physical space in the online world.

Together with the Zoom call and Twitch stream, nobody needed to colocate, and the productivity was better than on a traditional campus.

Christopher BrooksAssistant Professor at the University of Michigan
Christopher Brooks's avatar

After first learning about Deepnote from Professor Brooks,

... I started using the platform for my other projects, too. For example, it was an incredible help for analysing taxi data and weather data in another group.

Oleg NikolskyStudent of Master of Applied Data Science at the University of Michigan
Oleg Nikolsky's avatar


Overall, Deepnote allowed the group to talk and work as a team and reference things live in the stream while enabling everyone to work individually as needed. As a result, his group grew from Michigan to Hong Kong and other places in the world. Week by week, Chris has been successfully sharing his invaluable know-how in data science, a skill that his students cannot learn without the hands-on experience that they get.

Outcome for University of Michigan

Join the world's best data teams and get started with Deepnote

No credit card required. Run your first notebook in seconds.

© 2022 Deepnote. All rights reserved.