A Backend Roadmap Read From Job Postings
Where You Stand Now Decides the Order
In one line
What to do first depends on where you stand now. The paths are split into three.
Why is the order a problem?
There are 104 courses. "Just do all of them" is an answer that is useless to everyone — each course takes three or four hours, so doing everything takes months, and the posting closes in the meantime.
And if you get the order wrong, you spend time redoing what you already know while the real gaps stay as they are. So first decide where you stand now.
If you already work on a Python backend
The gaps are usually environment and scale.
- Production backend API capstone — it overlaps almost exactly with the first four lines of the posting. Eight steps, from schema design to ADRs and incident reports.
- Container basics and CKAD — fill in "experience using container technology" with labs.
- CI/CD pipelines and GitOps and ArgoCD — the posting names ArgoCD explicitly.
- Practical SQL — this is the preferred qualification "SQL performance optimization." Reading EXPLAIN comes before adding more indexes.
If Node.js is your main stack
The qualification says "Python or Node.js," so Node alone is enough to qualify. Even so, having the Python-side vocabulary keeps the conversation from stalling in the interview.
- Node.js backend — what frameworks hide — build a routing table, the output shape, and dependency injection yourself with the standard library.
- FastAPI — types are the contract — see what the same decisions are called on the Python side. If you know that Nest's pipes and FastAPI's response_model solve the same problem, you can follow the discussion in either direction.
- After that, follow the same path as above.
If you are a new graduate or junior
This posting asks for 3–5 years. The direction is still the same; only the sequence gets longer.
- Linux basics → Networking basics
- Database concepts → Practical SQL
- Docker basics → KCNA
- Then the backend capstone.
In the field — how to leave evidence that you prepared
The posting mentions "documentation" twice. Then the output of your preparation should be documents too.
- One ADR. What you chose, what you discarded, and what it cost. Step 7 of the capstone grades this.
- One incident report. One thing that actually blocked you during a lab is enough. Write it as impact, timeline, cause, detection, and correction.
- One number. Instead of "it was slow, so I fixed it," say "p95 was 1.4 seconds, and adding an index brought it to 210ms." That single sentence proves the preferred qualification of SQL optimization.
Three pages are enough. One document that explains a single decision all the way through goes further than a résumé listing fourteen tool names.