About
Hi, I’m Mario, and I have been running data science and bioinformatics trainings since 2013. These courses are technical and hands-on, with a lot of typing on the Linux command line. When some workshops moved online, I began looking for tools that could work alongside web conferencing platforms like Zoom.
I knew that the usual lab setup (shared instructions, local installs, VMs, Slack, and "please share your screen") would be brittle - I wanted participants to have zero local setup. I then looked at heavier platform approaches like Kubernetes-backed Jupyter environments. There were commercial offerings, as well as the option of setting up such an environment myself. But this would be:
- Difficult to set up: Off-the-shelf solutions provide only indirect access to the underlying machines, making it challenging to create and adapt a suitable training environment. And rolling your own Kubernetes-based setup is an over-the-top effort for most trainings.
- Costly: Spinning up many compute instances with above-average RAM or CPU becomes expensive quickly.
- Disconnected from participants: When participants get stuck, trainers can’t notice it early - and helping is slow, clunky, and disruptive.
That left me wishing for a lightweight solution that lets me better see where participants struggle and help them immediately when they run into trouble. ShellRoom was created to address these gaps and make live, terminal-first trainings smoother for everyone involved: instructors can monitor progress, chat, and instantly attach to a learner’s terminal - without screen-sharing chaos or heavyweight IDE setup.