FutureStackDev AI Agent Development

Software Engineering Fundamentals

Testing, Git in a team, reviews, APIs, and a deployment you can roll back. The habits AI features fail without.

Created by Baljeet Dogra

Course Objectives

  • Ship a small service with tests, a review, and a rollback path.
  • Read a stack trace and a failing CI job without panic.
  • Design a simple HTTP API and document the contract.
  • Treat prompts and models as dependencies of a normal service.

What you will produce

Service

Deployed API

A small Python service with tests, CI, and a container you can roll back.

Habit

Review notes

You sit as reviewer on a peer change and write what you would block on.

6 weeks curriculum

Expand a part for the syllabus. Content stays searchable when closed.

01 Git and the team Week 1

Branching, pull requests, bisect, and what never belongs in a commit.

  • Branching
  • PRs
  • History
  • Secrets
02 Testing that pays rent Week 2

Unit, integration, fixtures. What to mock. Coverage as a tool, not a trophy.

  • pytest
  • Fixtures
  • Mocks
  • CI
03 Reviews and design Week 3

How to leave a review someone can act on. When a design note beats another PR comment.

  • Review craft
  • ADRs in miniature
  • Interfaces
04 HTTP APIs Weeks 4–5

REST, status codes, idempotency, auth headers, versioning. The model is just another client.

  • Contracts
  • Errors
  • Idempotency
  • Auth
05 Ship and roll back Week 6

Docker, health checks, logs, a rollback drill. Capstone: the service off your laptop.

  • Containers
  • Health
  • Logs
  • Rollback

Who this is for

Scientists becoming engineers

You can train a model. You have not had to own a service.

Self-taught Python developers

You ship scripts. You want the habits a team will trust.

Prerequisites

  • Comfortable Python
  • A GitHub or GitLab account
  • You have run a program from the command line

Not a fit if

  • Staff engineers looking for distributed-systems depth
  • A management overview of “the SDLC”

Related: Python for GenAI · Data Structures & Algorithms

Questions

Will we use Kubernetes?

No. Containers and a rollback are enough here. Kubernetes for AI Systems is a later course.

Do I need to have deployed before?

No. By week 6 you will have.

Ready to start?

Six weeks to a service you can test, review and roll back. Create an account to enrol.

Enrol now