0–5 min: learner snapshot
Write the learner’s current coding level, math comfort, weekly time, and target role.
Success checkThe path is selected for a real learner, not an abstract persona.
Sequencing guide · Updated 2026-06-27
A practical sequencing guide for learners and teams choosing when to study web foundations, broad AI literacy, from-scratch neural networks, modern full-stack apps, and build-first AI systems.
The hardest learning-path decision is usually not which curriculum is best. It is when to use each curriculum. A beginner who jumps into neural-network derivations too early can stall; a developer who stays in broad AI overviews too long may never build; a team that assigns every path at once creates tutorial hoarding instead of capability.
This guide turns the RepoDaily Learning Radar into a sequence: start with the learner’s current skill, choose one primary artifact, then add the next curriculum only when the previous milestone has produced a reviewable project, notebook, or explanation.
RepoDaily verdict
Start with freeCodeCamp or The Odin Project for broad web foundations, use Full Stack Open when the learner is ready for rigorous modern full-stack apps, use Microsoft AI for Beginners to build broad AI literacy, use nn-zero-to-hero for neural-network mechanics, and use AI Engineering From Scratch when the learner can build and evaluate AI systems beyond notebooks.
| Stage | Best first path | Primary artifact | Move on when |
|---|---|---|---|
| 0. Programming and web basics | freeCodeCamp | Completed module plus one small deployed page or app | The learner can explain HTML/CSS/JS basics and debug without copying blindly |
| 1. Project-first web habits | The Odin Project | Portfolio project with Git history, README, and deployment | The learner can research docs, use Git, and finish a project without step-by-step guidance |
| 2. Modern full-stack app depth | Full Stack Open | React + backend app with tests and API/data notes | The learner can connect frontend, backend, tests, and deployment-like constraints |
| 3. Broad AI literacy | Microsoft AI for Beginners | Concept map plus quiz/assignment evidence | The learner can explain AI/ML/NLP/vision/responsible-AI vocabulary and choose a deeper path |
| 4. Neural-network mechanics | nn-zero-to-hero | Modified micrograd/makemore notebook plus written debugging note | The learner can explain gradients, loss curves, and model behavior in their own words |
| 5. Build-first AI systems | AI Engineering From Scratch | Small AI system or evaluation capstone with model/data/deployment notes | The learner can justify data, evaluation, inference, and maintenance tradeoffs |
Use this scorecard to choose the next curriculum, not to rank curricula universally.
| Path | Best learner state | Build artifact | Needs mentor review | AI depth | Web depth | Main risk |
|---|---|---|---|---|---|---|
| freeCodeCamp | Beginner needing structure | Exercises, certifications, small projects | Medium | Low | Medium | Breadth can become wandering |
| The Odin Project | Self-directed beginner-to-builder | Portfolio projects with Git history | High | High | High | Learners can stall without feedback |
| Full Stack Open | Post-beginner web developer | Modern full-stack app with tests | High | Low | High | Pacing can be intense |
| Microsoft AI for Beginners | AI beginner needing map | Concept map, quizzes, assignments | Medium | Medium | Low | Overview without implementation |
| nn-zero-to-hero | Developer ready for mechanics | Modified notebooks and derivations | High | High | Low | Too hard if math/debugging is weak |
| AI Engineering From Scratch | Developer ready to build AI systems | AI system capstone and evaluation notes | High | High | Low | Not a general beginner path |
Use this plan before assigning a learner or team a curriculum path.
Write the learner’s current coding level, math comfort, weekly time, and target role.
Success checkThe path is selected for a real learner, not an abstract persona.
Choose the first artifact: deployed page, portfolio app, full-stack app, concept map, modified notebook, or AI system capstone.
Success checkThere is one artifact that proves progress.
Open the candidate curriculum and inspect the first milestone plus one mid-course assignment.
Success checkThe difficulty is challenging but not blocking.
Name who reviews the artifact and what rubric they use.
Success checkFeedback is scheduled before the learner starts.
Define the condition that unlocks the next curriculum.
Success checkThe learner knows when to continue, switch, or pause.
| Scenario | Sequence | Why |
|---|---|---|
| Complete beginner wants to become a web developer | freeCodeCamp → The Odin Project → Full Stack Open | Start structured, build project habits, then add rigorous modern full-stack depth. |
| Junior developer wants to move into AI engineering | Microsoft AI for Beginners → nn-zero-to-hero → AI Engineering From Scratch | Build the AI map, then mechanics, then systems implementation. |
| Frontend developer wants backend/full-stack confidence | Full Stack Open → one internal capstone | The gap is app integration, tests, APIs, and production-like constraints, not another beginner path. |
| Product engineer needs AI literacy for reviews | Microsoft AI for Beginners → selected AI Engineering From Scratch labs | Broad vocabulary should come before implementation review responsibility. |
| ML-curious developer wants to stop treating models as black boxes | nn-zero-to-hero → AI Engineering From Scratch | Mechanistic notebooks prepare the learner for broader AI-system tradeoffs. |
| Team onboarding junior engineers | One web path + internal capstone + review rubric | The sequence must end in a project reviewed by the team, not a pile of links. |
| AI-assisted organization wants internal learning standards | Primary curriculum + AI-use contract + review checklist | AI tools should accelerate learning without hiding whether the learner understands the code. |
Collecting curricula feels productive but delays the first reviewed artifact. Lock the first four weeks to one primary path.
nn-zero-to-hero and AI Engineering From Scratch are powerful but punishing when assigned before coding and debugging habits exist.
Broad AI literacy must lead to a project, notebook, or review task; otherwise it becomes vocabulary without capability.
Learners need feedback on code, notebooks, explanations, and decisions. Completion alone is a weak signal.
AI coding tools can generate code faster than learners can explain it. Require manual explanation for submitted work.
Full Stack Open or The Odin Project should be adapted to the team’s stack only after the original learning objective is clear.
Use one curriculum as the spine and one supplement as context. More than that usually creates switching cost.
Do not move to the next path until the learner produces a project, notebook modification, concept map, or reviewed PR.
Run each stage as a four-week sprint with one weekly milestone and one review session.
Route learners by the capstone they need: portfolio app, full-stack app, AI concept map, neural-network notebook, or AI system.
Require learners to record where AI helped, what they verified manually, and what they can explain without assistance.
For teams, wrap external curricula with install steps, code review rubrics, expected artifacts, and examples from real work.
Short answers for teams sequencing AI and web learning paths.
If the learner cannot yet build, debug, and explain small programs, start with web/programming foundations first. AI courses become much more useful after basic coding confidence exists.
Use it after the learner can code and wants neural-network mechanics. It is a mechanics path, not a broad beginner AI overview.
Choose it when learners can already code and the goal is to build, evaluate, and reason about AI systems rather than only understand vocabulary.
Use one primary path and one small supplement. More than that usually creates switching cost and weak completion.
Feedback
Anonymous feedback helps RepoDaily improve what is actually useful.