Skip to main content
Version: Current

Core Systems

Learn the engineering, modeling, and deployment skills needed to build AI applications that others can run, evaluate, and use.

Start with Software Engineering for AI. Complete the five modules below and choose two electives.

Before you start​

Complete Engineering Fundamentals or demonstrate equivalent skills through the Academy direct-entry review. You should be able to write and debug Python, use a local environment and Git, manipulate data with NumPy/Pandas, query SQL, interpret plots, and use basic linear algebra, probability, and statistics. No prior FastAPI, Docker, PyTorch, or LLM application experience is required.

Your learning path​

How to work through the stage​

Start Software Engineering for AI and keep using its habits throughout the stage. Complete Applied ML before deep learning. Systems can run alongside modeling work. Begin APIs and deployment after systems and a working inference function from ML or deep learning.

Required exercises may share a repository with clear directories and run instructions.

Electives​

Complete exactly two electives after completing their prerequisites. Use each one to practice a defined capability. Further reading within a required module does not count as an elective.

Choose based on the capabilities you want to practice. Examples: LLM Applications and RAG + Agent Development for grounded workflows; Model Adaptation + MLOps for experimentation and delivery; Backend + MLOps for reliability. Other combinations are also valid.

Electives introduce capabilities explored more deeply in specialization. Keep each to one defined feature, an evaluation approach, and an explanation of tradeoffs. Both electives may share a repository, but document their outcomes and evidence separately.

Finish this stage​

Complete all five required modules and two electives. Resource completion supports progress; your implementations, explanations, debugging, and revisions demonstrate the learning outcomes. Watching tutorials alone does not establish them.

  • Correctness: your exercises behave as documented.
  • Evaluation: results use suitable baselines, splits/test cases, and metrics.
  • Engineering: another person can run the code, tests, and main workflow.
  • Understanding: you can explain decisions and diagnose a relevant failure.
  • Iteration: review feedback leads to a documented improvement.

Using AI during the stage. Use AI tools to explore, implement, test, and review. You are responsible for the final behavior and for explaining submitted work. Record one representative AI-assisted change and how you verified it; keep generated claims about results aligned with what your tests and evaluations actually show.

Resource access and scope. Use the assigned subset of each resource. Clean Code is further reading when access is available; the required engineering work can use the linked free documentation. Keep compute and deployment small, using local models, Academy-provided endpoints, or agreed infrastructure where suitable.

Continue into specialization afterward.

Plan your study time​

Use the provisional estimates where supplied. Other workloads remain pending pilot. Include implementation, debugging, review, and revision.