Roadmap

The course is divided into six learnable stages.

Use this page to track your milestones, review key concepts, and catch the classic Erlang misunderstandings early.

Roadmap

The course is divided into six learnable stages.

Overall progress

0%

0 of 16 learning steps completed

01

Think in values

Atoms, tuples, lists, pattern matching, and the mindset behind immutable data.

02

Think in functions

Guards, recursion, and case-based logic instead of classic loops.

03

Think in processes

Lightweight processes, mailboxes, messages, and asynchronous collaboration.

04

Think in resilience

Links, monitors, supervisors, and the logic behind let it crash.

05

Think in patterns

Workers, pipelines, state isolation, and common Erlang architecture patterns.

06

Think in systems

Nodes, distribution, deployment, and why BEAM is so strong in production.

Retention Lab

Come back to the core ideas until they feel obvious.

Real learning happens when you can recall the idea without looking. Use these cards as fast review rounds.

Why is immutable state so helpful in Erlang?

Because processes do not fight over shared mutable state, reasoning and concurrency get much simpler.

What is the mental model behind recursion?

Shrink the problem, keep the rule clear, and stop on a simple base case.

Why does message passing matter so much?

Because isolated processes coordinate by messages instead of hidden shared memory changes.

What does OTP really give you?

A standard structure for stateful processes, supervision, and recoverable systems.

Mistake Radar

These are the traps that make Erlang feel harder than it is.

Thinking in loops first

Beginners often search for loop syntax. In Erlang, think in recursive shape transformations instead.

Treating processes like objects

A process is not just a class instance. It is an isolated actor with a mailbox and explicit protocol.

Avoiding crashes at all costs

In BEAM systems, some failures should crash fast so supervision can restore clean state predictably.

Starting distributed design too early

First get one node right, then think about distribution, timeouts, retries, and monitoring.

Account

Login, register, and save your progress.

Account

Login, register, and save your progress.