Systems

Google Cloud Platform (GCP)

Strongest in data analytics (BigQuery), machine learning (Vertex AI), and Kubernetes (GKE — Google invented Kubernetes). Firebase is the best serverless backend for mobile apps. Pricing is competitive and sustained-use discounts apply automatically. The console UX is cleaner than AWS but the ecosystem is smaller. Best for: data-heavy workloads, ML pipelines, Firebase-backed apps, and teams that value developer experience over breadth of services.

Amazon Web Services (AWS)

The oldest and largest cloud platform. Over 200 services. If a cloud service exists, AWS has a version of it. EC2 (compute), S3 (storage), Lambda (serverless), RDS (managed databases), and DynamoDB (NoSQL) are the workhorses. Pricing is complex — reserved instances, spot instances, savings plans, and data transfer costs can surprise you. The console is sprawling. Best for: enterprises, teams that need every possible service, and workloads that benefit from the largest ecosystem of third-party integrations.

Microsoft Azure

The enterprise cloud. Strongest when you already live in the Microsoft ecosystem: Active Directory, Office 365, .NET, SQL Server, Visual Studio. Azure DevOps is a solid CI/CD platform. Azure OpenAI Service gives you GPT models with enterprise compliance baked in. Pricing is similar to AWS in complexity. Best for: enterprises with existing Microsoft contracts, .NET shops, and organizations that need Azure AD integration.

Claude / Anthropic

Anthropic builds the Claude family of models. Claude excels at long-context reasoning, code generation, nuanced instruction-following, and tasks that require careful analysis over large documents. The API is clean and well-documented. Claude Code is a CLI-native coding agent. The emphasis on safety research means the models tend to be more cautious and less prone to hallucination on factual claims. Best for: code generation, long-document analysis, agentic workflows, and teams that care about alignment.

OpenAI

The company that started the current AI wave with GPT-3 and ChatGPT. GPT-4o and o1/o3 are the flagship models. The API ecosystem is the largest — most third-party tools integrate with OpenAI first. DALL-E handles image generation, Whisper handles speech-to-text. ChatGPT Plus is the consumer product. Best for: general-purpose LLM tasks, image generation, speech processing, and when you need the broadest third-party integration support.

Local Models (Ollama)

Ollama lets you run open-weight models (Llama 3, Mistral, Phi, Gemma, DeepSeek) locally on your own hardware. No API costs, no data leaving your machine, no rate limits. The tradeoff is capability — local models are smaller and less capable than frontier cloud models. But for code completion, summarization, and lightweight chat, they are surprisingly good. Ollama runs on macOS, Linux, and Windows. If you have an M-series Mac with 16GB+ RAM, you can run 7B-13B parameter models comfortably. Best for: privacy-sensitive work, offline development, experimentation, and cost-conscious developers.

macOS

The dominant OS in frontend, mobile (iOS/macOS development requires a Mac), and design-heavy development. Unix-based, so the terminal is first-class. Homebrew is the de facto package manager. The hardware-software integration is unmatched — trackpad, display, battery life, and sleep/wake are simply better than anything else. Xcode is macOS-only. The downside: expensive hardware, less control over the system, and Apple ecosystem lock-in.

Linux (Ubuntu / Debian)

The server OS. Most production systems run Linux. Ubuntu is the most popular desktop distribution for developers — good hardware support, large community, apt package management. Debian is the stable, conservative choice that Ubuntu is built on. Linux gives you total control: you own your system down to the kernel. The downside: hardware support can be inconsistent (especially for laptops), and desktop polish lags behind macOS and Windows. Best for: backend developers, DevOps engineers, anyone who wants their dev environment to match production.

Windows

WSL2 (Windows Subsystem for Linux) changed the game. You can now run a full Linux kernel inside Windows with near-native performance. This means Windows is viable for almost any development workflow. Visual Studio (the full IDE, not VS Code) is still the best C#/.NET development environment. PowerShell is powerful but different from bash. Gaming and enterprise IT are Windows-dominated. The downside: the terminal experience is still rougher than macOS/Linux, and some open-source tools treat Windows as a second-class citizen.

Mac Hardware (MacBook Pro / Mac Studio)

Apple Silicon (M1 through M4 Ultra) is the most power-efficient chip architecture available. A MacBook Pro with 18+ hours of battery life that compiles code faster than most desktops is not a luxury — it is a competitive advantage. Unified memory means the GPU and CPU share the same RAM pool, which is excellent for ML inference (Ollama runs beautifully on M-series Macs). The 14" and 16" MacBook Pro are the most popular developer laptops in the industry. Mac Studio is the desktop powerhouse for video editing, 3D rendering, and heavy compilation.

Lenovo ThinkPad

The developer laptop for people who do not use Macs. The ThinkPad X1 Carbon and T-series have been the gold standard for Linux laptop compatibility for over a decade. Legendary keyboards — many developers swear the ThinkPad keyboard is the best ever put in a laptop. Excellent build quality, business-grade reliability, and easy repairability. The X1 Carbon is ultraportable; the T-series is the workhorse. Best for: Linux developers, enterprise environments, and anyone who values keyboard quality and repairability over display and trackpad refinement.

iPhone / iOS Devices

If you build iOS apps, you need physical iOS devices for testing. The simulator covers most cases, but push notifications, camera, Bluetooth, haptics, and performance profiling require real hardware. An iPhone and an iPad cover the two major form factors. TestFlight handles beta distribution. The App Store review process is strict but well-documented. Beyond development: the iPhone is the most popular smartphone among developers, largely because of iMessage, AirDrop, and Handoff integration with macOS.

Android Devices

Android development requires testing across a wider range of devices because of hardware fragmentation. A recent Pixel phone is the best single test device — it runs stock Android and gets updates first. Samsung Galaxy devices are important for market coverage since Samsung dominates Android market share. The Android emulator in Android Studio is excellent but slow on machines with limited RAM. USB debugging and wireless ADB are your daily drivers. Budget devices matter: your app needs to perform well on a 3-year-old phone with 3GB of RAM, not just on the latest flagship.

VS Code

The most popular code editor in the world by a wide margin. Free, open-source core (Code - OSS), runs on every platform, and has the largest extension ecosystem. GitHub Copilot, ESLint, Prettier, GitLens, and language-specific extensions make it viable for almost any language. The integrated terminal, built-in Git support, and remote development (SSH, containers, WSL) set the bar. It is not the fastest editor and it is Electron-based (so it uses more memory than native apps), but the ecosystem is unbeatable.

Terminal / iTerm2

The terminal is the most powerful tool most developers underuse. On macOS, iTerm2 is the gold standard: split panes, search, profiles, triggers, and shell integration. On Linux, GNOME Terminal, Alacritty (GPU-accelerated), and Kitty are popular. Pair your terminal with a good shell (zsh with Oh My Zsh, or fish), a multiplexer (tmux), and a prompt (Starship). The developers who are fastest are almost always the ones who live in the terminal.

Xcode

Apple's IDE for iOS, macOS, watchOS, tvOS, and visionOS development. It is the only option — you cannot build and sign iOS apps without it. Interface Builder and SwiftUI previews are integrated. Instruments (the profiler) is powerful but complex. Xcode is notorious for large download sizes, slow indexing, and occasional instability. But it is also the most vertically integrated mobile development environment available: code, design, test, profile, sign, and submit from one app.

Android Studio

Google's IDE for Android development, built on JetBrains IntelliJ IDEA. Gradle-based build system, layout editor, device emulator, and profiling tools are all integrated. Kotlin is the primary language (Java still works). The emulator is resource-hungry — budget 8GB+ of RAM just for the IDE and emulator together. Flutter development also works well in Android Studio via the Flutter plugin, though VS Code is more popular for Flutter.

JetBrains IDEs

IntelliJ IDEA (Java/Kotlin), PyCharm (Python), WebStorm (JavaScript/TypeScript), GoLand (Go), CLion (C/C++), and Rider (.NET). The JetBrains suite offers the deepest language-specific intelligence: refactoring, navigation, and code analysis that VS Code extensions approximate but rarely match. They are paid products (except IntelliJ Community Edition). Heavy on system resources. Best for: developers who work deeply in a single language and want the best possible tooling for that language rather than a general-purpose editor.

Figma

The industry standard for UI/UX design. Browser-based, real-time collaboration, powerful component system, auto-layout, and a massive plugin ecosystem. Dev Mode generates CSS, iOS, and Android code snippets from designs. The free tier is generous. Figma acquired by Adobe (pending regulatory approval as of 2024, then abandoned). Best for: any team doing UI design. It has effectively won the design tool war for product design.

Sketch

The original modern UI design tool. macOS-only. Still used by teams that adopted it before Figma and have large existing design systems in Sketch format. Symbols and libraries are powerful. The plugin ecosystem is mature. The downside: no real-time collaboration (Sketch for Teams exists but lags behind Figma), no browser version, and the broader industry has moved to Figma. Best for: macOS-only teams with legacy Sketch files.

Adobe Creative Cloud

Photoshop (raster), Illustrator (vector), After Effects (motion), Premiere Pro (video), InDesign (print). Adobe dominates professional creative work outside of UI design. Photoshop is still the best raster editor. Illustrator is still the best vector editor for print and illustration work. The subscription model is expensive ($55/month for the full suite). Adobe XD (their Figma competitor) has been effectively discontinued. Best for: print design, photo editing, motion graphics, and video production.

Canva

The design tool for non-designers. Template-driven, drag-and-drop, browser-based. Excellent for social media graphics, presentations, simple videos, and marketing collateral. The AI features (Magic Design, background removal, text-to-image) are surprisingly capable. Not suitable for detailed UI/UX work or pixel-precise design. Best for: marketing teams, solo founders, content creators, and anyone who needs good-enough design fast.

CapCut

Free video editor from ByteDance (TikTok's parent company). Available on desktop and mobile. Surprisingly powerful for a free tool: multi-track timeline, keyframe animation, auto-captions, green screen, speed ramping, and a large library of effects and templates. The auto-caption feature alone makes it worth using — it generates accurate subtitles with styleable text. Best for: short-form content, social media videos, quick edits, and anyone who needs auto-captions without paying for a subscription.

DaVinci Resolve

The most powerful free video editor available. Used by Hollywood colorists (the color grading tools are industry-leading). The free version includes a full NLE (nonlinear editor), Fusion (visual effects and motion graphics), Fairlight (audio post-production), and the legendary color page. The Studio version ($295 one-time, not subscription) adds GPU acceleration, HDR tools, and collaboration features. The learning curve is steep but the ceiling is effectively unlimited. Best for: serious video production, color grading, anyone who refuses to pay monthly for editing software.

OBS Studio

Open-source software for live streaming and screen recording. The standard tool for Twitch streamers, YouTube live, and developer screencasts. Scene composition, audio mixing, multiple sources (screen capture, camera, window capture, browser sources), and plugin support. Free, cross-platform, and battle-tested by millions of streamers. The UX is not pretty but it is functional and reliable. Best for: streaming, screen recording, virtual camera setups, and live presentations.

Final Cut Pro

Apple's professional video editor. macOS and iPad only. Magnetic timeline (love it or hate it — it prevents gaps and collisions), excellent performance on Apple Silicon, and tight integration with the Apple ecosystem (AirDrop footage from iPhone, edit on Mac). The $300 one-time purchase is a significant cost advantage over Adobe Premiere Pro's subscription. ProRes support is native and fast. Best for: Mac users who want professional editing without a subscription, and creators who shoot on iPhone or iPad.

PostgreSQL

The most capable open-source relational database. Full SQL compliance, JSONB support (store JSON documents with indexing — blurring the SQL/NoSQL line), full-text search, PostGIS (geospatial), and excellent extension ecosystem (pg_vector for AI embeddings, TimescaleDB for time-series). If you are choosing one database and have no other constraints, choose Postgres. It handles more use cases well than any other single database.

MySQL

The other major open-source relational database. Simpler than Postgres, faster for simple read-heavy workloads, and has the largest installed base (WordPress, most PHP applications, older web apps). InnoDB is the default storage engine and handles transactions well. The MySQL vs Postgres debate is mostly settled: Postgres is more capable, but MySQL is adequate for many workloads and has more hosting options at the low end.

MongoDB

The most popular document database. Stores data as JSON-like documents (BSON). No schema enforcement by default (though you can add validation). The query language is JavaScript-based. Atlas (the managed cloud service) is well-executed. MongoDB excels when your data is naturally document-shaped (content management, catalogs, user profiles) and when schema flexibility matters more than relational integrity. The downside: no joins (you denormalize instead), and data consistency guarantees are weaker than SQL databases by default.

Firebase / Firestore

Google's serverless database platform. Firestore is a document database with real-time listeners — your UI updates automatically when data changes. Authentication, hosting, cloud functions, and storage are bundled. The free tier is generous (Spark plan). Firestore scales automatically and requires zero database administration. The tradeoff: vendor lock-in (your data model is shaped by Firestore's query limitations), limited query capabilities (no full-text search, no joins, inequality filters on one field only), and costs can spike unpredictably at scale. Best for: mobile apps, prototypes, real-time applications, and teams that want to eliminate backend infrastructure entirely.

Redis

An in-memory data store used primarily as a cache, message broker, and session store. Sub-millisecond read/write latency. Data structures beyond simple key-value: lists, sets, sorted sets, hashes, streams, and pub/sub. Redis is not a primary database (data lives in memory, persistence is optional and has tradeoffs). It sits in front of your primary database to cache hot data and reduce load. Also used for rate limiting, leaderboards, real-time analytics, and job queues (via Sidekiq, BullMQ, etc.).

SQLite

A serverless, file-based SQL database embedded directly in your application. No separate server process — the database is a single file on disk. Used by more applications than any other database engine (every iPhone, every Android phone, every browser, every macOS machine has SQLite). Excellent for: mobile apps (the default database on iOS and Android), desktop apps, CLI tools, embedded systems, and development/testing (use SQLite locally, Postgres in production). The limitation: single-writer concurrency. If you need multiple processes writing simultaneously, use a client-server database.

JavaScript / Node.js

The language of the web. Every browser runs JavaScript. Node.js brought it to the server. The ecosystem (npm) is the largest package registry in the world. JavaScript is dynamically typed, single-threaded (with an event loop for async), and extremely flexible — for better and worse. Its ubiquity is its greatest strength: one language for frontend, backend, mobile (React Native), desktop (Electron), and serverless. Its greatest weakness: the same flexibility that makes it productive also makes it error-prone at scale, which is why TypeScript exists.

TypeScript

JavaScript with static types. Created by Microsoft, now used by the majority of professional JavaScript projects. TypeScript catches entire categories of bugs at compile time that JavaScript only catches at runtime (or never). The type system is one of the most expressive in any mainstream language — union types, intersection types, conditional types, template literal types, and mapped types let you model complex domains precisely. The learning curve beyond basic types is steep. Best for: any JavaScript project larger than a script. The upfront cost of type annotations pays for itself immediately in editor intelligence and refactoring safety.

Python

The most popular language for AI/ML, data science, scripting, and automation. Readable syntax that reads almost like pseudocode. The ecosystem is vast: NumPy, pandas, scikit-learn, TensorFlow, PyTorch, FastAPI, Django. Python is slow compared to compiled languages, but that rarely matters because the hot path is usually in C/C++ libraries (NumPy operations run in compiled C, not interpreted Python). Best for: machine learning, data analysis, scripting, backend APIs, and teaching programming. Not ideal for: mobile apps, frontend web, or performance-critical systems.

Go

Google's language for building reliable, efficient, concurrent systems. Statically typed, compiled to native binaries, garbage collected, and opinionated about code formatting (gofmt). Goroutines and channels make concurrent programming approachable. The standard library is excellent — you can build an HTTP server, parse JSON, and write tests without any third-party dependencies. Go deliberately omits features (no generics until 1.18, no exceptions, no inheritance) to keep the language simple. Best for: backend services, CLI tools, DevOps tooling (Docker, Kubernetes, Terraform are all written in Go), and systems that need to handle high concurrency.

Swift

Apple's language for iOS, macOS, watchOS, tvOS, and visionOS development. Modern, safe (optionals eliminate null pointer exceptions), fast (compiled, with performance approaching C++), and expressive (protocol-oriented programming, value types, powerful enums with associated values). SwiftUI is the declarative UI framework. Swift is effectively required for Apple platform development — Objective-C still works but new APIs are Swift-first. The language evolves quickly (Swift concurrency with async/await and actors landed in Swift 5.5). Best for: anything on Apple platforms.

Dart

Google's language, primarily used with Flutter for cross-platform mobile, web, and desktop development. Dart compiles to native ARM code (mobile), JavaScript (web), and native x64 (desktop). The language is clean, familiar (similar to Java/TypeScript), and has excellent async support (Futures and Streams). Flutter's hot reload makes development fast. The ecosystem is smaller than JavaScript or Python, and outside of Flutter, Dart has limited adoption. Best for: cross-platform app development with Flutter.

Rust

A systems programming language focused on safety, speed, and concurrency. The borrow checker enforces memory safety at compile time without a garbage collector — no null pointer dereferences, no data races, no use-after-free. The learning curve is the steepest of any mainstream language (the borrow checker will fight you until you internalize the ownership model). Once you get past the curve, Rust code is remarkably reliable. Performance is comparable to C/C++. Best for: systems programming, WebAssembly, performance-critical services, CLI tools, and anywhere you need the performance of C with the safety of a modern language. Not ideal for: rapid prototyping or teams that need to move fast with junior developers.

Java / Kotlin

Java is the enterprise backbone — banks, insurance companies, large-scale backend systems. The JVM (Java Virtual Machine) is one of the most optimized runtimes ever built. Spring Boot is the dominant backend framework. Java is verbose but reliable, well-understood, and has a massive talent pool. Kotlin is a modern JVM language (created by JetBrains) that is fully interoperable with Java but more concise, safer (null safety built in), and more expressive. Kotlin is the preferred language for Android development (Google made it official in 2019). Best for: enterprise backends (Java/Spring), Android development (Kotlin), and any project that benefits from the JVM ecosystem.

The best tech explainer on the internet. His "100 seconds" format distills complex technologies into sharp, fast, visual overviews. The longer videos are equally good — clear, opinionated, and information-dense. If you watch one tech YouTuber, make it Fireship.

Former Netflix senior engineer. Deep dives into algorithms, data structures, Vim/Neovim workflows, and systems programming. Entertaining, opinionated, and technically rigorous. His "Frontend Masters" courses on algorithms are excellent. If you want to get better at the fundamentals, watch Prime.

Former Twitch engineer, creator of the T3 Stack (Next.js, tRPC, Tailwind, Prisma). Strong opinions on TypeScript, React, and the JavaScript ecosystem. Fast-paced commentary on new releases, drama, and industry trends. Best for: staying current on the TypeScript/React/Next.js ecosystem.

Networking, cybersecurity, Linux, and cloud infrastructure explained with high energy and beginner-friendly production. His series on Docker, Kubernetes, and home labs are great entry points for developers who want to understand infrastructure. Best for: developers moving into DevOps or wanting to understand networking fundamentals.

Former Google/Facebook tech lead. Covers career advice, technical interviews, Silicon Valley culture, and the business side of software engineering. Dry humor and contrarian takes. Best for: career navigation and understanding how big tech companies operate from the inside.

Creator of Linux and Git. Does not produce content in the traditional sense, but his talks (especially the TED talk and various Linux Foundation keynotes) and mailing list posts are masterclasses in systems thinking, technical decision-making, and opinionated engineering leadership.

Definitive articles on software architecture, design patterns, refactoring, and engineering practices. Dense, long-form, and authoritative. His writing on microservices, event sourcing, CQRS, and continuous delivery are industry references. Read slowly and revisit often.

Former Tesla AI director, OpenAI founding member. His "Neural Networks: Zero to Hero" YouTube series is the best free resource for understanding how large language models work from first principles. If you want to understand AI beyond the API level, Karpathy is the teacher.

Full-stack developer content with a focus on React, GraphQL, TypeScript, and developer lifestyle. His older technical content is solid. Known for an irreverent, meme-heavy style that makes dense topics approachable.

Rich Hickey's "Simple Made Easy" (the most important talk on complexity). Bryan Cantrill's talks on debugging and systems programming. Dan Abramov's React talks. Kelsey Hightower's Kubernetes demos. Gary Bernhardt's "Wat" and "The Birth & Death of JavaScript." These are evergreen — watch them regardless of when they were recorded.

freeCodeCamp

Free, self-paced, project-based curriculum covering HTML/CSS, JavaScript, Python, data structures, APIs, and more. Over 10,000 hours of content. The certifications are not prestigious in the credentialing sense, but the projects you build along the way are real portfolio pieces. The YouTube channel is also excellent — full-length courses on everything from Docker to machine learning. Best for: beginners and career changers who need structured, free, hands-on learning.

The Odin Project

A free, open-source full-stack curriculum. Two paths: Ruby on Rails and JavaScript/Node.js. The key differentiator is that The Odin Project does not hold your hand — it teaches you to read documentation, debug independently, and build projects from scratch. The community (Discord) is active and supportive. Best for: self-motivated learners who want to build real full-stack skills, not just follow along with tutorials.

Udemy

Massive marketplace of video courses on every technology. Quality varies wildly — some courses are outstanding, others are outdated cash grabs. The trick: never pay full price (courses go on sale for $10-15 regularly), check reviews and publication dates, and look for instructors with multiple highly-rated courses. Notable instructors: Colt Steele (web dev), Stephen Grider (React/Node), Maximilian Schwarzmuller (Angular/React/Flutter), Jose Portilla (Python/data science). Best for: visual learners who want structured video walkthroughs of specific technologies.

Coursera

University-level courses from Stanford, MIT, Google, and other institutions. The specializations (multi-course sequences) are the best value. Andrew Ng's Machine Learning course is legendary — it launched thousands of ML careers. Google's IT Support and Data Analytics certificates are practical and employer-recognized. Courses are free to audit; certificates require payment ($39-79/month). Best for: developers who want academic rigor and recognized credentials.

MIT OpenCourseWare

Free access to MIT's actual course materials: lectures, problem sets, exams, and solutions. The computer science courses (6.006 Introduction to Algorithms, 6.824 Distributed Systems, 6.828 Operating Systems) are world-class. These are not simplified tutorials — they are real MIT courses. The pace and difficulty reflect that. Best for: developers who want deep theoretical foundations and are comfortable with academic-level material. Pair with YouTube lecture recordings for the full experience.

Frontend Masters

Premium video courses taught by industry practitioners (not career instructors). ThePrimeagen teaches algorithms, Kyle Simpson teaches JavaScript (You Don't Know JS), Scott Moss teaches Node.js, Steve Kinney teaches React performance. The production quality is high and the depth is aimed at intermediate-to-advanced developers. $39/month. Best for: working developers who want to go deeper on specific technologies with expert instructors.

Exercism

Free coding practice platform with mentored tracks in 60+ languages. You solve exercises locally in your own editor, submit solutions, and optionally receive feedback from human mentors. The value is practicing a new language through progressively harder exercises with idiomatic feedback. Best for: learning a new programming language by writing code, not watching videos.