8 Tech Roles, 4 Levels Each: The Free Learning Map

Pick a role, climb four levels, build one project at each. No paid courses required.

Author

Vishakha Sadhwani
November 04, 2026

Every tech role looks overwhelming from the outside. Open any job description and you'll find twenty tools, a dozen buzzwords, and "5+ years of experience" stapled to the end.

Strip it down, though, and most roles stand on just four layers. Learn them in order and each one makes the next easier: the language first, then the platform, then the practice, then the thing that makes the role special.

Below are eight roles, the four levels for each, and 100% free resources to learn every level, including the crash courses. Pick one role, go top to bottom, and build something small after every level.

1. Forward Deployed Engineer

Python → APIs → RAG → Agents

What the job is: You sit with a customer, understand their messy real-world problem, and ship an AI system that solves it inside their stack. Half engineer, half problem-solver on site.

Level 1: Python

Level 2: APIs

Level 3: RAG

Level 4: Agents

Put it together

2. DevOps Engineer

SDLC → Linux → CI/CD → Containers

What the job is: You build the road code travels on, from a developer's commit to a running service, and you make that road fast, safe, and boring.

Level 1: SDLC

Level 2: Linux

Level 3: CI/CD

Level 4: Containers

Put it together

3. Cloud Engineer

Compute → Networking → Security → Multi-Cloud

What the job is: You design and run the infrastructure everything else sits on, and you're the person who knows why the bill doubled last month.

Level 1: Compute

Level 2: Networking

Level 3: Security

Level 4: Multi-Cloud

Put it together

4. Data Engineer

SQL → ETL → Spark → Vector Search

What the job is: You move data from where it's created to where it's useful, clean and on time. Today that includes feeding AI systems, not just dashboards.

Level 1: SQL

Level 2: ETL

Level 3: Spark

Level 4: Vector Search

Put it together

5. Robotics Engineer

Python → C++ → ROS 2 → Isaac Sim

What the job is: You write the software that lets a machine sense, decide, and move, and you test it in simulation long before it touches real hardware.

Level 1: Python

Level 2: C++

  • LearnCpp.com, the best free C++ course on the internet, start to finish

Level 3: ROS 2

Level 4: Isaac Sim

Heads up: Isaac Sim needs an NVIDIA RTX GPU. No GPU? NVIDIA's learning paths point to cloud GPU options, or start with ROS 2's built-in tools and Gazebo.

6. Machine Learning Engineer

Python → Statistics → PyTorch → CUDA

What the job is: You take models from notebook to production, which means understanding the math well enough to debug them and the hardware well enough to make them fast.

Level 1: Python

Level 2: Statistics

Level 3: PyTorch

Level 4: CUDA

Put it together

7. Site Reliability Engineer

Linux → Monitoring → SLOs → Reliability

What the job is: You keep production up, decide how "up" is up enough, and turn every outage into a fix that stops it from happening again.

Level 1: Linux

Level 2: Monitoring

Level 3: SLOs

Level 4: Reliability

Put it together: Take any project from the DevOps section, add a Prometheus dashboard, write one SLO for it, then break it on purpose and write the postmortem.

8. Software Development Engineer

Python → DSA → System Design → Distributed Systems

What the job is: You build the product itself, and you're expected to write code that's correct, fast, and still works when ten million people use it.

Level 1: Python

Level 2: DSA

Level 3: System Design

Level 4: Distributed Systems

Crash Course Quick List

Every crash course from this article in one place.

Foundations

DevOps and Containers

Cloud

Observability

AI

KodeKloud Crash Courses

How to Use This

Don't try to learn all eight. Pick the role that excites you, and treat its four levels as a ladder: don't move up until you've built one small thing with the level you're on. A Python script, an API, a RAG bot over your notes, an agent that calls that API. Four levels, four projects, and suddenly the job description doesn't look so scary.

You'll also notice how much overlaps. Python shows up in four roles, Linux in three, APIs and Kubernetes everywhere. Nothing you learn here is wasted if you switch lanes later.

All links are free courses, docs, books, labs, or crash courses. Course dates and availability change, so if a link has moved, search the title and you'll usually find the current version.