Relearning a technology by rebuilding it exposes assumptions textbooks often hide. This collection gathers concise, well-scoped “build your own …” guides so you can reconstruct the internals of systems (from a tiny OS to an LLM) and learn the design decisions engineers actually make.
What Sets It Apart
- Curated, cross-language tutorials with concrete targets (e.g., a ray-tracer, a tiny Redis, an LLM-from-scratch). So what: you get minimal, implementable projects instead of broad surveys—good for deliberate practice.
- Wide coverage that bridges systems and AI (neural nets, diffusion models, RAG, vision pipelines). So what: it’s a single index when you want to connect low-level engineering with ML/AI concepts.
- Community-maintained links and contributions with clear entry points per topic. So what: new tutorials and language variants appear via PRs, keeping the list practical and varied.
- Lightweight learning-by-doing focus—examples and blogs over heavy libraries. So what: you’ll understand core algorithms and trade-offs rather than only learning an API.
Who It's For and Trade-offs
Great fit if you learn best by implementing: students, engineers preparing interviews, or practitioners who want to demystify black-box tools by rebuilding core components. Look elsewhere if you need production-ready libraries, step-by-step tooling for deploying large-scale ML, or a single cohesive tutorial path—this repo is an index of many independent guides, so depth and quality vary by entry.
Where It Fits
Use this as a hands-on syllabus: pick a target implementation (e.g., a search engine or a tiny LLM) and follow one or more linked guides to build intuition. For formal coursework or managed ML workflows, complement these tutorials with textbooks or platform-specific docs (e.g., PyTorch docs, MLops guides).
