What engineers say after completing the programmes

"The mathematics bridge was the right thing for me, though I underestimated the pace during the probability section. Problem sets are not easy — which is the point. Worked solutions after submission are genuinely useful because you can see exactly where your reasoning diverged from the correct approach. Office hours saved me twice."

"I joined the deployment course after inheriting a model at work that I did not fully understand how to serve. The load-testing exercise was the most practically useful thing I have done in a course in years. Twelve hours a week is not an exaggeration — plan for it. The incident simulation is uncomfortable in a good way."

"The residency was a different experience from any course I have taken. Having a single project to carry from start to finish over twenty weeks is more demanding than I expected. The external design reviews are not formalities — reviewers ask hard questions about design decisions. My main feedback is that the data governance module could be slightly earlier in the sequence."

"I tried two other online AI courses before this one and both assumed I already had the mathematics. The bridge course was the first that explained why linear algebra matters for training, not just how it works in isolation. The code implementations alongside every concept made it stick."

"The deployment course has a specific hardware ladder that tells you upfront what you can do on a laptop versus the cluster. That kind of transparency is unusual. The cost and carbon accounting session is something I have not seen covered anywhere else — useful given how quickly inference costs add up in production."

"I am partway through the residency. The dedicated mentor is genuinely useful — I have a specific person who has read my design documents and can respond to the actual decisions I made, not generic advice. The in-person week involved more peer review than I expected, which was harder and more productive than lectures would have been."