The Zero to Elite roadmap: 21 phases from beginner to job-ready engineer

Zero to Elite is a phase-based software engineering curriculum of about 3,034 estimated hours, roughly 17.5 months at 40 hours a week. Each phase explains why it comes where it does, what it depends on, and what you should be able to show before moving on. Follow it with the ZTE Tracker app to track every topic, DSA problem and project.

Foundation phases

Tools, language fundamentals, React and interface systems. 4 phases, about 613 hours.

  1. Phase 01: Developer Environment & Foundations

    About 185 hours

    Everything later runs in a terminal, an editor and a Git repository, and the web runs on HTML, CSS and JavaScript. Time lost to a broken environment or to not understanding the language underneath is paid back many times over.

    What you should be able to show

    • A repository with several meaningful commits you can explain
    • A static multi-page site built without a framework, deployed publicly
    • A short JavaScript program you can debug with devtools and a failing test
  2. Phase 01b: TypeScript Mastery

    About 80 hours

    TypeScript turns many runtime bugs into editor errors and is the working language of most modern web codebases. It comes right after JavaScript so frameworks are learned with types from the start, not retrofitted.

    What you should be able to show

    • A small module with generics and a discriminated union, with tests
    • A tsconfig you can explain option by option
    • A runtime validation schema paired with the inferred static type
  3. Phase 02: React + Next.js Core

    About 213 hours

    Components, state and routing are how interfaces are built today. This phase comes after JavaScript and TypeScript so learners understand what the framework does for them.

    What you should be able to show

    • A multi-route app with server and client components used deliberately
    • Forms with validation and visible loading and error states
    • A deployed app with a README explaining the data flow
  4. Phase 03: UI System & Styling

    About 135 hours

    Consistent, accessible, responsive interfaces need a system of tokens and components rather than one-off CSS. It follows core React so components exist to be styled.

    What you should be able to show

    • A themed component set with a documented API
    • An automated accessibility check passing on the main journey
    • A layout that holds up from a phone to a wide desktop screen

Core phases

State, backend, databases, testing, deployment and DSA. 6 phases, about 1,135 hours.

  1. Phase 04: State, Data Fetching & Advanced React

    About 153 hours

    Real applications spend most of their complexity in server state, caching and asynchronous behaviour. This comes after the React basics so the cost of each pattern is understood.

    What you should be able to show

    • An app with cached, invalidated server state and optimistic updates handled safely
    • A reproduced and fixed race condition in an asynchronous flow
    • A written note on where state lives and why
  2. Phase 05: Backend + Database + Auth

    About 314 hours

    A product needs durable data, authenticated users and an API. Backend, database and authentication are grouped because they only make sense together, and they follow the frontend so learners have something real to connect.

    What you should be able to show

    • An authenticated API with a relational schema, migrations and a tested authorisation rule
    • Stable pagination and an idempotent create endpoint with tests
    • A query plan before and after an index change
  3. Phase 06: Testing + CI/CD + Deployment

    About 155 hours

    Code that is not tested and deployed through a pipeline is a demo. This phase turns earlier projects into software a team could maintain, so it sits right after the first full-stack system exists.

    What you should be able to show

    • CI that runs type checks, tests and linting on every change
    • Tests at unit, integration and end-to-end level for one feature, with the reasoning for each level
    • A deployment with a documented rollback
  4. Phase 06b: React Native (Mobile)

    About 111 hours

    Mobile reuses React knowledge but adds lifecycle, offline behaviour and store release. It is optional and placed after web fundamentals and testing so it builds on them.

    What you should be able to show

    • An app that keeps state across a force-quit
    • Offline writes that sync without duplicates
    • A build on a beta track with crash reporting
  5. Phase 07: API Documentation & Developer Tooling

    About 71 hours

    An API other people can use needs a contract, documentation and tooling. It follows the backend phase because there must be an API to describe.

    What you should be able to show

    • An OpenAPI description that matches the implementation and is checked in CI
    • A documented error format used consistently
    • A client or collection that exercises the documented examples
  6. Phase 08: DSA & Interview Preparation

    About 331 hours

    Algorithmic problem solving is tested in many interviews and sharpens reasoning about performance. It runs alongside project work from early on, since skill grows from steady practice, and sits here as a deliberate block once the language is solid.

    What you should be able to show

    • A log of solved problems with the pattern, the complexity and a mistake you made on each
    • Timed solutions to unfamiliar problems explained aloud
    • A note of the patterns you still miss, with a plan

Advanced phases

Caching, monitoring, real-time, AI and production hardening. 7 phases, about 842 hours.

  1. Phase 09: Caching + Email + Payments

    About 98 hours

    Caching, email and payments are the first dependencies on external systems that fail in new ways. They follow the backend and testing phases so failure handling can be tested.

    What you should be able to show

    • Cached responses with a stated invalidation rule
    • Webhook handling that is safe when the same event arrives twice
    • Email sending that is retried without duplicates
  2. Phase 10: Monitoring & Analytics

    About 103 hours

    You cannot operate or improve what you cannot see. Logs, metrics and traces turn production behaviour into evidence, and the phase follows deployment because there must be something running to observe.

    What you should be able to show

    • Structured logs with a request identifier across services
    • A dashboard and one alert tied to user impact
    • A real or simulated incident investigated from the data
  3. Phase 11: Real-Time + Search + PostgreSQL Internals

    About 164 hours

    Real-time features, search and database internals are where simple designs strain. They follow the first backend and observability work so learners can measure the strain.

    What you should be able to show

    • A real-time feature that survives reconnects
    • Search with measured relevance on a small labelled set
    • A slow query diagnosed and fixed from its plan
  4. Phase 12: AI/RAG + Production AI Patterns

    About 207 hours

    AI features are now common in products, and building them safely needs retrieval, evaluation and defences against misuse. They sit after backend, testing and observability, since an AI feature is a service that needs all three.

    What you should be able to show

    • A retrieval feature with recall measured on labelled questions
    • A golden evaluation set run as a regression test
    • A documented set of defences against prompt injection and the tools the model may call
  5. Phase 13: Advanced Browser APIs + Collaboration

    About 137 hours

    Advanced browser capabilities and collaboration patterns extend frontend skill once the basics, state and testing are strong.

    What you should be able to show

    • A feature using an advanced browser API with graceful fallback
    • A collaboration pattern with a stated conflict rule
    • A measured performance comparison before and after
  6. Phase 14: Load Testing + Security Deep Dive

    About 89 hours

    Load behaviour and security flaws appear only when an application is real. This phase comes after a deployed, observable system exists to test and harden.

    What you should be able to show

    • A load test with a stated budget and the bottleneck found
    • Injection, XSS and access-control flaws reproduced on a practice app and fixed with tests
    • A threat model for one of your own projects
  7. Phase 15: Build Tooling + CSS-in-JS

    About 44 hours

    Build tooling and CSS-in-JS explain what happens between source and browser. Studying them later means learners already have a project whose build they want to understand.

    What you should be able to show

    • A bundle analysis with one change that reduced size, measured
    • A build configuration you can explain
    • A trade-off note on a styling approach

Expert phases

Infrastructure, architecture, craft and career. 4 phases, about 444 hours.

  1. Phase 16: Infrastructure

    About 175 hours

    Infrastructure turns code into a running, repeatable environment. It follows deployment and observability because those show why repeatability and safe change matter.

    What you should be able to show

    • An environment created and changed only through code with reviewed plans
    • A small, non-root container image
    • A rollout with a rollback you exercised
  2. Phase 17: Architectural Patterns

    About 178 hours

    Architectural patterns are trade-offs between simplicity and the ability to change. They are best studied after a few systems have been built, so the problems the patterns solve have been felt.

    What you should be able to show

    • A written design with alternatives and the reason each was rejected
    • A refactor that applies one pattern, with tests proving behaviour did not change
    • A failure analysis of your own design
  3. Phase 18: Engineering Craft + Package Publishing

    About 62 hours

    Engineering craft, such as code review, documentation, refactoring and publishing a package, is what makes work maintainable by others. It is practised against the codebases built in earlier phases.

    What you should be able to show

    • A published package or an accepted contribution to someone else's project
    • A pull request with a description a reviewer can act on
    • A refactor with a before and after explanation
  4. Phase 19: Career & Community

    About 29 hours

    Evidence only helps if it reaches an employer. This phase turns the work of earlier phases into a resume, a portfolio, interview practice and an honest job-search routine.

    What you should be able to show

    • A resume where every bullet links to evidence you can show
    • Two projects with READMEs a stranger can run
    • Recorded mock interviews with notes on what you will change

Ship the roadmap, one topic at a time

Create an account to track progress, lock topics until their prerequisites are done, and schedule revision.