$ yuktics v0.1

A complete CS curriculum · v0.1

The CS curriculum,
rebuilt for now.

What to study, what to build, what stack to know — and how to use AI without becoming dependent on it. From your first line of code to your first negotiated offer. The replacement for the old GeeksForGeeks-era playbook, for students who started a CS journey after the world changed.

Start here Pick a learning path
Tracks
8
Modules
44
Ready now
44
License
MIT

Why this exists

Five years ago a CS student could grind GeeksForGeeks, CLRS, and a couple of MOOCs and feel ready. That playbook is now half-broken. Memorizing every algorithm matters less when an AI can produce one in seconds — but the students who can't reason about systems, can't debug, and can't evaluate AI output are getting filtered out faster than ever. This guide is what that fork in the road needed: the old foundations, kept where they still pay; the new skills, slotted in where they actually belong; and the AI layer treated as a power tool, not a religion.

Every module follows the same shape

  1. 01 Outline — what you'll be able to do at the end.
  2. 02 Stack — exact tools, libraries, models.
  3. 03 Resources — papers, repos, posts ranked by signal.
  4. 04 Build — a concrete project with milestones.
  5. 05 Checkpoints — questions you should answer cold.

Pick a starting path

Same curriculum, different entry points. Pick the row that sounds most like you and start there.

High schooler / no code yet

You've never written a line. You're not even sure if CS is for you.

  1. 00.1
  2. 00.2
  3. 00.3
  4. 01.1
  5. 01.2
  6. 03.1
  7. 01.3
  8. 02.1

By the end of this path: you've shipped a website, you can use a terminal without fear, and you know whether CS is your thing.

CS freshman / sophomore

You're in a degree, you know basic Python, you want to pull ahead of your cohort.

  1. 00.3
  2. 00.4
  3. 01.2
  4. 01.3
  5. 01.4
  6. 02.1
  7. 02.2
  8. 02.6
  9. 03.2
  10. 03.3
  11. 06.1

By the end of this path: you have a real full-stack project on GitHub, fluency in the modern web, and CS theory grounded in code you wrote.

Pivoting toward AI

You can already code. You want to be employable in AI within a year.

  1. 00.3
  2. 04.1
  3. 04.2
  4. 04.5
  5. 04.6
  6. 04.7
  7. 05.1
  8. 06.3

By the end of this path: you've built a transformer and a real agent, and you can speak technically about deploying both.

Late-degree / interviewing soon

You're 12–18 months from job hunting. You want to compress.

  1. 02.2
  2. 02.7
  3. 02.8
  4. 05.1
  5. 05.2
  6. 05.4
  7. 07.1
  8. 07.2
  9. 07.4

By the end of this path: you can pass a system-design interview, negotiate your offer, and walk in on day one without imposter syndrome.

Full roadmap

Eight tracks, 44 modules. Take them in order, follow a learning path, or pick the next module that closes the gap you actually have. Read freely. Steal liberally.

  1. T0

    The Meta Layer

    How to learn CS in 2026. The advantage you have that students five years ago didn't.

    1. 00.1 How to learn CS in the AI era ready Why the old playbook (memorize syntax, grind tutorials, read textbooks cover-to-cover) is now obsolete — and what replaces it. ~2 hrs prereq: curiosity
    2. 00.2 Your dev environment, properly set up ready macOS / Linux / WSL setup, terminal, dotfiles, package managers, an editor that doesn't fight you. ~3 hrs prereq: a laptop
    3. 00.3 Using AI as your tutor ready Claude, ChatGPT, Cursor, Claude Code. When to ask, when not to, how to learn from the answer instead of copying it. ~3 hrs prereq: 00.2
    4. 00.4 Reading code, reading docs, reading errors ready The three skills that compound the most. None of them are taught in school. ~2 hrs prereq: 00.2
  2. T1

    Programming Foundations

    One language deeply, then a typed one, then a systems one. In that order.

    1. 01.1 Python deep enough to be dangerous ready Functions, classes, comprehensions, generators, async, packaging. Skip the parts that don't matter. ~10–14 hrs prereq: 00.2
    2. 01.2 The terminal, shell, and Unix ready bash/zsh, pipes, grep/awk/sed, find, processes, ssh, tmux. The skill that pays compound interest forever. ~5–7 hrs prereq: 00.2
    3. 01.3 Git and GitHub, properly ready Commits, branches, rebase vs merge, PRs, conflict resolution. Stop fearing your VCS. ~4–6 hrs prereq: 01.2
    4. 01.4 TypeScript: a typed language that pays ready Types, generics, narrowing, the React/Node ecosystem. The most-employable language in 2026. ~8–12 hrs prereq: 01.1
    5. 01.5 C or Rust: closer to the metal ready Pointers, memory, ownership. You only need this once, but you need it once. ~12–18 hrs prereq: 01.1
  3. T2

    CS Theory You Actually Need

    The textbook half of CS, taught by building. Algorithms, OS, networks, databases, architecture.

    1. 02.1 Data structures, built not memorized ready Build the dozen that matter (array, linked list, stack, queue, hash map, tree, heap, graph) once each, by hand. Then never re-implement them. ~10–14 hrs prereq: 01.1
    2. 02.2 Algorithms & complexity, the practical view ready Big-O without the academic framing. The 12 algorithm patterns that solve 80% of interview problems. ~12–16 hrs prereq: 02.1
    3. 02.3 Discrete math + the math behind algorithms ready Counting, recursion, induction, graphs, modular arithmetic. The minimum that lets you read CS papers. ~8–12 hrs prereq: high-school math
    4. 02.4 Operating systems: the parts that matter ready Processes, threads, memory, files, syscalls. Build a tiny shell. Understand what `top` is showing you. ~12–18 hrs prereq: 01.5
    5. 02.5 Computer networks, end to end ready What happens when you type a URL? TCP, DNS, HTTP, TLS — built up from sockets. ~10–14 hrs prereq: 02.4
    6. 02.6 Databases & SQL, deep ready PostgreSQL as your default. Indexes, transactions, joins that don't melt servers, when to use NoSQL. ~10–14 hrs prereq: 01.1
    7. 02.7 Computer architecture & performance ready Cache, memory hierarchy, branch prediction, why your code is slow. Profiling as a first-class skill. ~8–12 hrs prereq: 01.5
    8. 02.8 Distributed systems, intro ready Why your laptop's mental model breaks. Replication, consensus, the eight fallacies, real-world failures. ~8–12 hrs prereq: 02.5, 02.6
  4. T3

    Build Things People See

    Web, frontend, backend, deployment. Ship a real product before you graduate.

    1. 03.1 Web fundamentals without the cargo cult ready HTML, CSS, JavaScript. The ~30% of each you actually need. Skip the bootcamp pile-on. ~8–12 hrs prereq: 01.4
    2. 03.2 Modern frontend: React + Next.js ready Components, state, routing, server components. Build a real app, not a counter. ~12–18 hrs prereq: 03.1
    3. 03.3 Modern backend: APIs, auth, databases ready FastAPI or Hono. REST, JSON, JWT, sessions, password hashing — without copying StackOverflow. ~10–14 hrs prereq: 02.6
    4. 03.4 PostgreSQL like a pro ready Schema design, migrations, indexes, EXPLAIN, full-text search, JSONB, the queries that break in production. ~8–12 hrs prereq: 02.6
    5. 03.5 Deploying & DevOps lite ready Docker, GitHub Actions, Vercel/Fly/Render. Putting your project on the internet so a recruiter can click it. ~6–10 hrs prereq: 03.3
    6. 03.6 Mobile: where it actually matters ready When to build a mobile app vs. a responsive site. React Native + Expo for the cases that need it. ~8–12 hrs prereq: 03.2
    7. 03.7 Ship a real full-stack project ready Spec → schema → API → UI → deploy → monitor. The capstone of T3. ~20–40 hrs prereq: 03.5
  5. T4

    AI Literacy & Engineering

    From using AI like a power user to building real AI systems.

    1. 04.1 The math you actually need for ML ready Linalg, probability, calculus, optimization — ranked by how often it shows up. Less than the textbook tells you. ~6–10 hrs prereq: 02.3
    2. 04.2 Transformers from scratch ready Build nanoGPT in plain PyTorch. Attention, training loop, sampling — the highest-leverage weekend in this guide. ~8–12 hrs prereq: 04.1
    3. 04.3 Tokenizers, datasets, training ready BPE, packing, masking, mixed precision, gradient accumulation. The unsexy work that makes models good. ~6–8 hrs prereq: 04.2
    4. 04.4 Fine-tuning: LoRA, QLoRA, full FT ready When to use which. Real configs that work. Adapter merging. ~5–8 hrs prereq: 04.3
    5. 04.5 Build an AI agent ready Tool use, agent loop, tracing, eval. Ship a real research agent in 200 lines, no framework. ~6–10 hrs prereq: 04.2
    6. 04.6 RAG done right ready Chunking, embeddings, hybrid search, rerankers, query rewriting. Why most RAG demos fall apart in production. ~6–10 hrs prereq: 04.2
    7. 04.7 Inference, deployment, costs ready vLLM, quantization, KV cache, batching. The economics of running your own models. ~5–8 hrs prereq: 04.4
  6. T5

    System Design & Scale

    Reasoning about systems bigger than one server. The interview track that's also real.

    1. 05.1 Designing systems on a whiteboard ready The interview format, but also the actual skill. Scoping, capacity, trade-offs, naming what you don't know. ~8–12 hrs prereq: 02.8
    2. 05.2 Caching, queues, rate limits ready Redis, Kafka, exponential backoff, idempotency keys. The standard kit of any production backend. ~6–10 hrs prereq: 03.3
    3. 05.3 Observability & ops ready Logs, metrics, traces. What to instrument, what alerts mean, what `99.9% uptime` actually allows. ~5–8 hrs prereq: 03.5
    4. 05.4 Security: the parts you can't skip ready OWASP Top 10, auth flows that don't leak, secrets management, threat modelling at the right level. ~6–10 hrs prereq: 03.3
  7. T6

    Build Your Reputation

    The part of being employable that nobody teaches in CS class.

    1. 06.1 Side projects that compound ready Most side projects die in week three. The ones that don't share a shape — and that shape is recognizable. ~2 hrs prereq: 03.5
    2. 06.2 Open source: from contributor to maintainer ready Finding good first issues, the etiquette of PRs, becoming someone the maintainer wants to merge from. ~ongoing prereq: 01.3
    3. 06.3 Technical writing & public proof ready Blog posts, READMEs, talks. How writing publicly is the most under-priced career move in tech. ~ongoing
    4. 06.4 Communities worth being in ready Discord, X, hackathons, local meetups. Where the actual hiring conversations happen. ~ongoing
  8. T7

    The Career

    Internships, interviews, first job, negotiation. The part where the curriculum meets payroll.

    1. 07.1 Internships: getting and using your first one ready When to apply, what gets read, what doesn't. The cold-email template that actually works. ~5 hrs prereq: T1+T2 in flight
    2. 07.2 The interview gauntlet ready LeetCode without becoming a robot. System design. Behavioral. The interviews that pretend to be technical but aren't. ~ongoing prereq: 02.2, 05.1
    3. 07.3 The first job: 90 days, 6 months, 2 years ready What to optimize for at each horizon. The mistakes new grads make on autopilot. ~3 hrs
    4. 07.4 Negotiating your offer ready The single highest hourly-rate conversation you will have for a decade. Don't wing it. ~3 hrs
    5. 07.5 Generalist or specialist (and the trap of both) ready How to decide what to commit to, when, and what to refuse to commit to even when it pays better. ~2 hrs