Learning by making
CommonLab.
Understanding grows when you get to try things.
The most interesting moment in a workshop is often when someone stops following a demonstration and starts asking their own question. CommonLab is an attempt to make more room for that moment.
A place to build a mental model
AI vocabulary can arrive faster than understanding. Models, agents, tools, memory, and workflows make more sense when a learner can assemble something, run it, inspect what happened, and change it. CommonLab turns that learning loop into a guided environment.
Learners build real agents in controlled sandboxes, using the tools and sample data an educator has made available. Courses and short skill paths provide structure; logs and outputs make behaviour inspectable. Educators can bring their own curriculum, with a console for courses, sessions, and practice environments.
Connect the tool to the work
The platform sits alongside the workshops and training I facilitate with organisations, including Moringa School and Tallinn University. In Nairobi, I worked with Jessica Francisca Colaço on AI Skills for Everyday Work: practical exercises for professionals with different technical backgrounds.
Those sessions give the learning work a grounding in everyday questions. What information is appropriate to share? How do we check an answer? Where does human review belong? My intended outcome is confidence backed by understanding: people who can build something useful and make thoughtful decisions about how it fits into their work.
Some doors into the project, its origins, and the people around it.