Скачать или смотреть Clean Architecture in Python — Boundaries, DDD, and Testing

Clean Architecture in Python focuses on drawing clear boundaries so code survives framework churn and product change. In this conversation with Sam Keen, author of Clean Architecture with Python, we cover practical techniques for keeping volatile details at the edges, protecting the domain from coupling, and sustaining testability as systems grow.

Sam outlines what “clean” looks like in a dynamic language: enforcing the Dependency Rule, using repositories and inversion to keep frameworks outside core logic, and making pragmatic validation choices with dataclasses or Pydantic—captured in ADRs when teams choose a specific path. We also discuss project structures that make good defaults easy, along with lightweight checks that catch boundary leaks early.

We dig into applied DDD in Python—entities and value objects—plus a testing strategy that emphasizes fast unit feedback with purposeful integration and end-to-end coverage. The session covers incremental legacy refactoring with strangler-fig patterns and bounded contexts, service boundaries and event hygiene, and the role of AI in real workflows through context engineering and grounded code generation.

Whether you’re a Python engineer, tech lead, or architect working across several services, this is a grounded discussion on making Python codebases more reliable and evolvable under change.

Main Topics Covered:

Dependency Rule and boundaries that outlast frameworks
Repositories and inversion to isolate frameworks
DDD in Python: entities, value objects, and invariants
Validation choices: Pydantic and dataclasses, captured in ADRs
Testing strategy: fast unit feedback; focused integration and end-to-end checks
Legacy refactoring: strangler-fig, bounded contexts, parity checks
Service boundaries and message design that avoid internal leakage
Clarifying core and edge responsibilities
AI’s role: context-first workflows and safe code generation

About the guest:
Sam Keen is a software engineering leader with over 25 years of experience. He’s a polyglot developer who has used Python across early-stage startups and large-scale systems at companies including AWS, Lululemon, and Nike. His current focus is developer experience—building tools and systems that amplify developer productivity through intelligent automation. As a principal engineer at Pluralsight, he sets foundational standards for distributed architecture while staying close to code and delivery, with a strong interest in making AI practically useful to engineering teams.

He authored- Clean Architecture with Python (Packt)

Packt: https://www.packtpub.com/en-us/product/cle...

Amazon: https://www.amazon.com/Clean-Architecture-...

Subscribe to Deep Engineering for more expert-led insights on software architecture, performance, and real-world development:
https://deepengineering.substack.com/