What participants say
about the cohorts
Feedback from developers who completed Promptly's five-week, fourteen-week, and thirty-two week programmes. No edited marketing quotes — just what people wrote in their post-cohort responses.
Back to HomeParticipant Reviews
"I had Python basics but no idea how to run models outside of notebook tutorials. The five-week track filled that gap well. The assignment on data preparation took me longer than expected — which was actually useful, since it surfaced how much I didn't know about structuring inputs. The code review feedback was specific and made me rewrite parts of my repository properly."
"The engineering practice cohort matched what I was looking for — I could already write Python but my project organisation was a mess. Going through version control and reproducibility requirements properly made a real difference to how I structure work now. The written summaries were a grind but the habit has carried over. I'd have liked slightly more time on evaluation, which felt compressed in the final weeks."
"Finished the full developer programme in January. The RAG systems section was the one I came for and it covered it at a depth I hadn't found elsewhere. Four sprints over thirty-two weeks is a lot of work when you're doing this part-time, and the workload was real — but it was accurately described upfront so I knew what I was getting into. Capstone review was thorough."
"Joined the engineering practice track after finding that most online courses didn't have anyone actually looking at my code. The difference is real. The reviewer caught a structural issue in my sprint 1 project that I would have dragged into sprint 2 otherwise. The written materials are dense but well-organised — searchable without needing to rewatch anything."
"The five-week cohort was the right entry point for me. I had done some Python for data work but hadn't tried running a model outside of Colab. Getting through local inference setup properly and then the API section was what I needed. The portfolio repo gave me something concrete to show for it. It's not a long programme, but the scope is honest about what it covers."
"Sprint 3 of the full developer programme — the small-team practices section — was something I hadn't seen covered in any other course I looked at. Having to write up project decisions and submit them for review, not just submit code, was different from what I'd done before. Took some adjustment. The final capstone documentation is something I've actually referenced back since finishing."
Participant Journeys
Challenge
A data analyst with two years of Python experience had used pre-built APIs but had never set up a local model environment or written code to call an open-source model directly. Most tutorials assumed either no Python knowledge or significant ML background — neither fitted.
Through the cohort
The five-week track provided structured guides on environment setup, local inference, and API use. Assignment 1 covered model calls with structured outputs; assignment 2 required preparing and feeding a small dataset through a pipeline. Code review identified dependency management issues in the first submission.
Outcome
Completed both assignments and the portfolio repository within the cohort window. The repository included two working pipeline examples with documented run instructions. The participant enrolled in the fourteen-week track in the following cohort intake.
"Getting the environment working properly in week one was actually the thing I'd been stuck on for months. Once that was sorted, the rest followed."
Challenge
A backend engineer had been running small AI experiments for personal projects but found them difficult to share or reproduce later. No consistent structure, no evaluation discipline, no way to hand them off to a colleague. The code worked but the engineering was informal.
Through the cohort
The fourteen-week programme covered version control patterns, model serving, evaluation frameworks, and reproducibility requirements. Three sprints required progressively more complete project documentation. Sprint 2 focused specifically on structuring evaluation pipelines — the area where the participant had the weakest foundation.
Outcome
All three sprints and the capstone submitted and reviewed. Post-cohort, the participant applied the same documentation and reproducibility habits to a project at their employer — a RAG system built with a small team — and reported that the structure they learned during the programme transferred directly.
"The sprint format forced me to actually finish things rather than leaving half-done notebooks everywhere. That's still how I work now."
Get in Touch
Sat: 10:00 AM – 1:00 PM
Ready to join a cohort?
Contact the team with your current Python background and which track interests you. Responses within one working day.
Request a Quote