Six engineering disciplines
You get hired for what you can show, run and defend. Every LEECOI track ends in a portfolio a hiring engineer can open, execute and ask you hard questions about.
Tooling from the published syllabi
The catalogue
Fundamentals, then applied building, then a portfolio someone can open and run. The discipline changes. The shape does not.
LLM application engineering. Retrieval, vector databases, agents, and the backend work that turns a demo into something a company can actually run.
Classical machine learning, end to end. You build tabular models the way industry actually builds them, then tune, interpret and document them well enough to defend in an interview.
Infrastructure as code, delivery pipelines that ship safely, and the observability that makes production legible.
Offensive and defensive security, practised in environments that behave like production rather than like an exercise.
Pipelines, warehouses, and the modelling that decides whether the numbers downstream can be trusted.
Full-stack web applications, from the data layer to the interface, built and then actually deployed.
How it goes
Progress is not linear, and pretending otherwise is how people quit in week nine believing they are the problem. The curve at the top of this page is the honest version, dips included.
01
The early weeks move quickly, because fundamentals are well-trodden and the feedback is immediate. Python, SQL, Git, the shell, and whichever mathematics the track leans on. This is the stretch that feels good.
02
Then it flattens. You are building real things against real messiness, and some weeks end further back than they started. This is the longest stretch, and getting through it is the whole difference between a portfolio and a certificate.
03
The final weeks are documentation, error analysis, and rehearsing the questions an interviewer will ask about your own code. Nothing new is introduced. Everything becomes defensible.
Phase boundaries differ by track: the Machine Learning Engineer syllabus runs five phases, the AI Engineer syllabus four. Each track page carries its own week-by-week roadmap.
After the building
Your profile is usually read before your code is. Both get the same review treatment.
Every project read the way a hiring engineer would read it, then sent back for another pass.
Structured practice on your own code, with feedback on how the answer landed rather than only on whether it was right.
The same review treatment, applied to the thing recruiters read before they read any code.
The search has its own skills. Treated as part of the programme, not as what happens after it.
Straight answers
Before you ask
No. The published tracks expect Python, SQL, Git and Linux basics, plus algebra, probability and statistics fundamentals. Where you learned them does not matter. Each track page lists its prerequisites in full, so you can check before committing.
A portfolio of repositories with documented problem statements, baseline and final results, error analysis and honest limitations. The Machine Learning Engineer track ends in two capstone projects, the AI Engineer track in three. All five briefs are on this page.
Two are. The Machine Learning Engineer and AI Engineer syllabi are written and published. The syllabi for Cloud Engineering and DevOps, Cybersecurity, Data Engineering and Analytics, and Web Development are still being written, and no invented curriculum stands in for them in the meantime. Registering interest is how you hear when one is ready.
No. There is no placement percentage, no salary figure and no job promise anywhere on this site, because none of those can be evidenced. What is offered is portfolio review, interview preparation, profile support and job search guidance, which are the parts within our control.
Register interest and we will send the full week-by-week syllabus for the track you are considering, then arrange a short call to work out whether it is the right one.
No payment details, and nothing to install. The syllabus is the thing being sent.