You can write code. The job post asks for everything else.
Docker. Kubernetes. CI/CD. APIs. AWS. I mentor engineers one-on-one until deploying and operating a real system is something you have actually done — not something you have read about.
The skills are not hard. The absence of a map is.
Courses teach you models and syntax. Job descriptions ask whether you can put something in front of real users and keep it alive. That second list is what filters people out of applications they were otherwise qualified for.
You stop guessing what words mean
Containers, orchestration, pipelines, infrastructure as code — explained in the order they actually matter, on your own code.
You ship, not just build
Something running in the cloud, tested, deployed automatically, with logs you can read when it breaks.
You apply without flinching
You read a job description and recognise the work, because you have done a smaller version of it yourself.
One project, carried all the way to production.
Everything is built around a single project that is yours. You pick the domain — an API, a model service, a data pipeline, whatever is closest to the job you want — and we take it from empty repository to something deployed, tested and observable.
At the end you are not holding a certificate. You are holding a URL, a repository and the ability to explain every decision in it.
-
01
Intro call — 20 minutes
Where you are, which jobs you are aiming at, and whether this mentorship is the right thing for you right now.
-
02
We choose the project
Scoped to your target role and small enough to finish, big enough to be worth showing.
-
03
90 minutes a week, live
Working sessions, not lectures: we build, debug and review together, and I explain the reasoning as we go.
-
04
Async support in between
Questions as you go? Send them on our channel.
-
05
You finish with a portfolio piece
Deployed, documented, and rehearsed — so you can walk an interviewer through it.
Who this is for, and who it isn't.
This is for you if
- You know some Python and can build things locally, but have never put one in production.
- You are a data scientist or analyst moving toward ML or software engineering.
- You skip job applications because of the infrastructure section of the description.
- You can commit around three hours a week for three to four months.
This is not for you if
- You want to study transformer internals or research-level ML theory.
- You have never programmed before — start with the fundamentals first.
- You are looking for interview question drills or a certificate to collect.
- You want someone to build the project for you.
Gustavo Tomsic
I work on production ML systems using Python, AWS, APIs, cloud infrastructure and modern engineering tooling. The mentorship is the path I wish someone had drawn for me: fewer tutorials, one real system, and the vocabulary to talk about it.
View LinkedInTwenty minutes to find out if this is your missing piece.
Fill this in and you go straight to my calendar to pick a time.