Curriculum overview

The program runs for 12 weeks in three phases. Each phase builds on the one before, and everything you learn feeds into your capstone: a real product, deployed and running by week 12.

The curriculum follows the skills a working forward deployed engineer relies on, based on the roadmap.sh FDE roadmap (opens in a new tab): software foundations, backend and frontend development, AI engineering, production skills and customer delivery. You'll deploy something live in weeks 3, 5 and 10, so shipping becomes a habit long before the capstone.

This page is the map. Each week's overview goes into the detail. Before week 1, install the tools in the setup guide.

Phase 1 · Foundations

Weeks 1–3. You learn to write Python that does real work, store data and put your first service on the internet. By the end of this phase, you have a small API of your own running live.

Week 1 · Python and your developer toolkit

• Using the terminal to move around and run programs
• Tracking your work with Git and sharing it on GitHub
• Working efficiently in your code editor
• The Python you need to write small, useful scripts

Week 2 · Python for real work and SQL

• Organising code into functions and modules
• Reading and writing files and JSON
• Keeping projects separate with virtual environments
• Handling errors so programs fail clearly instead of crashing
• Storing and querying data with SQL and SQLite

Week 3 · The web, APIs and your first deployment

• HTTP requests and responses, URLs and status codes
• Calling JSON APIs from Python
• Building a small backend API with FastAPI
• Deploying it to the cloud so anyone can reach it

Phase 2 · Building

Weeks 4–8. You build AI-powered products: backend, frontend and the AI features in between. By the end of this phase, you can build a deployed AI application that answers from real data and connects to the tools a customer already uses.

Week 4 · Calling LLM APIs

• Sending requests to a large language model
• Writing prompts that give consistent results
• Getting structured output back instead of free text
• Handling real-world failures such as timeouts and bad responses
• Measuring accuracy with precision, recall and a confusion matrix

Week 5 · Frontend and a deployed AI app

• HTML and CSS for a clean, simple interface
• JavaScript that calls your own backend API
• Keeping API keys on the server, never in the browser
• Deploying the full app with a link you can share
• Stretch: rebuilding the interface in React

Week 6 · Retrieval (RAG)

• Why and where organisations use retrieval
• Splitting documents into chunks and turning them into embeddings
• Vector search and retrieval pipelines
• Answers that cite their sources

Week 7 · Making retrieval reliable

• Improving results with reranking and metadata filters
• Diagnosing why retrieval fails
• Evaluating answer quality with repeatable tests (evals)

Week 8 · Agents and automation

• Choosing between plain automation, an LLM step and an agent
• Tool use, function calling, state and memory
• Retries, fallbacks, guardrails and human approval steps
• Connecting to business systems with APIs, webhooks and n8n

Phase 3 · Delivering

Weeks 9–12. You work like an FDE: scope a real problem, build the solution, get it into production and present it. By the end of this phase, you have a deployed, monitored capstone and the experience of defending it.

Week 9 · Working with customers

• Discovery: understanding a customer's problem before building
• Scoping what to build and what to leave out
• Working with stakeholders
• Writing a business case: the problem, the users and the value
• Your capstone proposal

Week 10 · Production and reliability

• Packaging your capstone in a Docker container
• Deploying to the cloud with proper configuration and secrets
• Retries, rate limits and graceful failure
• Logging and monitoring usage, errors, speed and cost

Week 11 · Capstone build

• Building and integrating your capstone end to end
• Evaluating its quality with evidence
• Architecture diagram and documentation
• Capstone checkpoint

Week 12 · Demo day

• Final polish and deployment checks
• Portfolio, GitHub profile and a short demo video
• Presenting your capstone live
• Technical validation: explaining, changing and debugging your own code

Throughout the program

Some skills aren't taught in a single week because you use them all the time:

• Git and GitHub. Every assignment lives in a repository, so version control becomes a habit.
• AI coding assistants. You'll learn to use tools like Claude Code and Cursor to work faster, and to check what they produce. See the AI tools policy for the ground rules.
• Writing and communication. Explaining your work clearly is part of the job, in READMEs, demos and updates.
• Security and privacy. Handling secrets, user data and access safely comes up in every phase, not just at deployment.