Prompting & Context
Writing instructions a model can follow: goals, constraints, examples, and the right context in the window.
The foundations every developer still needs.
AI writes the surface. Learn what is underneath, plus the new skills to direct, review and trust what AI builds.
New
Skills for the AI era
Work that exists because AI now writes code.
Reviewing AI-Written Code
Read generated code like a pull request from a confident junior: check logic, edge cases, and what it silently assumed.
Debugging Code You Didn't Write
Form a hypothesis, narrow it down, and fix the real cause instead of re-prompting until it stops failing.
LLM APIs & Tool Use
Calling models from code: messages, streaming, structured output, function calling, and handling failures and cost.
Evals & Hallucination Checks
Measure whether an AI feature works: test sets, graders, and ways to catch confident wrong answers.
AI Security
Prompt injection, leaked secrets, over-permissioned agents, and untrusted model output used as code.
Core
Core fundamentals
What every developer needs, and why it matters more now.
Binary & Number Systems
Binary, hexadecimal, bitwise operations. Computers speak in 0s and 1s — understanding that changes how you think.
In the AI eraExplains why models quantize weights and why floats lose precision.
Calculus for ML
Derivatives, gradients, chain rule. Backpropagation is just calculus — understanding it makes you a better AI engineer.
In the AI eraGradients are how models learn; the math hasn't changed.
Design Patterns
Proven solutions to recurring problems. Patterns make code readable, scalable, and maintainable.
In the AI eraA shared vocabulary to direct AI and review what it builds.
How Computers Work
Memory, CPU, processes, threads, the call stack. What actually happens when you run your code.
In the AI eraTells you why a model call is slow, costly, and memory-hungry.
Version Control
Git isn't optional. Branching, merging, resolving conflicts — this is how real teams build software.
In the AI eraYour safety net when an AI edit rewrites 40 files at once.
Networking Basics
HTTP, DNS, TCP/IP, APIs, request/response cycles — the internet is your runtime.
In the AI eraEvery LLM call is an HTTP request: timeouts, retries, streaming.
Linear Algebra
Vectors, matrices, dot products, transformations. The language every AI model speaks under the hood.
In the AI eraEmbeddings and attention are matrix operations.
Big O Complexity
Understand how your code scales. Time and space complexity are what separates good code from great code.
In the AI eraSpot the quadratic loop AI wrote that passes the demo and dies in prod.
Programming Paradigms
Object-oriented, functional, procedural — knowing when and why to use each one is a core skill.
In the AI eraLets you judge whether generated code fits your codebase's style.
Data Structures
Arrays, linked lists, stacks, queues, trees, graphs — the building blocks every algorithm depends on.
In the AI eraChoosing the right structure is still your call, not the model's.
Algorithms
Sorting, searching, recursion, dynamic programming. Learn to think in solutions, not just code.
In the AI eraYou can't review a solution you couldn't have reasoned toward.
Databases
Relational vs non-relational, queries, indexing, and why data modeling matters from day one.
In the AI eraAI writes queries fast; you catch the missing index and the bad schema.
Security Basics
Encryption, authentication, common vulnerabilities. Every developer is responsible for writing secure code.
In the AI eraAI code ships leaked secrets and injection holes with full confidence.
Testing
Unit, integration, end-to-end. If you don't test it, you don't know it works.
In the AI eraThe only honest way to trust code you did not write.
Operating Systems
Processes, scheduling, file systems, permissions — the layer between your code and the hardware.
In the AI eraDebugging a hung process still needs you, not the chat window.
APIs & Protocols
REST, WebSockets, GraphQL, gRPC. How systems talk to each other is as important as the systems themselves.
In the AI eraTool use and agents are APIs talking to APIs.
Compilers & Interpreters
How source code becomes execution. Lexing, parsing, ASTs — what actually runs your program.
In the AI eraHelps you read the errors the model can't explain.
Package Management
npm, pip, cargo — managing dependencies, versioning, and avoiding supply chain chaos.
In the AI eraModels hallucinate package names; attackers register them.
Concurrency & Async
Threads, event loops, promises, async/await. Writing code that does more than one thing at a time is a skill in itself.
In the AI eraGenerated async code looks right and races anyway.
Statistics & Probability
Distributions, variance, Bayes theorem, hypothesis testing. ML is applied statistics — skip this and you're guessing.
In the AI eraThe base for evals: is this model really better, or just lucky?
Command Line
Navigate, automate, and control your environment. The terminal is where real development happens.
In the AI eraAgents run shell commands. You must know what they are doing.
Engineering
Beyond the code
System design, architecture and the habits that keep real software running.
System Design
Break a product into services, data stores and queues, and reason about load, latency and failure before you build.
In the AI eraAI drafts components fast; you decide the trade-offs it can't see.
Software Architecture
Layers, boundaries, monolith vs services, and how to keep a codebase changeable as it grows.
In the AI eraGenerated code piles up without a structure to hold it. You own the structure.
API Design
Resources, versioning, pagination, errors and idempotency: contracts other people build on.
In the AI eraModels copy popular API habits, including the bad ones.
Scalability & Caching
Horizontal scaling, load balancing, caches and where the real bottleneck is.
In the AI eraPremature scaling is easy to generate. Knowing what to measure first is not.
Clean Code & Refactoring
Naming, small functions, removing duplication and changing code safely under tests.
In the AI eraAI adds code faster than anyone deletes it. Refactoring keeps it readable.
Code Review
Give and take review: spot risk, ask the right questions and keep changes small.
In the AI eraEvery AI change is a pull request you must actually review.
Containers & CI/CD
Docker, build pipelines and automated deploys so shipping is boring and repeatable.
In the AI eraPipelines are your guard rail when changes arrive at AI speed.
Observability & Production Debugging
Logs, metrics and traces: find out what a live system is doing and why it broke.
In the AI eraAI can't see your production. Telemetry is how you find out what it did there.
Reliability & Failure Handling
Timeouts, retries, backoff, rate limits and graceful degradation when dependencies fail.
In the AI eraGenerated happy-path code rarely handles the failure paths.
Queues & Event-Driven Design
Background jobs, message queues and events: decouple work and survive spikes.
In the AI eraAsync pipelines multiply the bugs AI can't reproduce for you.
Technical Writing & Decisions
READMEs, design docs and architecture decision records that explain why, not just what.
In the AI eraAI writes the docs. You decide what is true and worth recording.
FAQ