Maintainable
Robby Russell
Do you feel like you're hitting a wall with your existing software projects? Are you curious to hear how other people are navigating this? You're not alone.
On the Maintainable Software Podcast, Robby speaks with seasoned practitioners who have overcome the technical and cultural problems often associated with software development.
Our guests will share stories in each episode and outline tangible, real-world approaches to software challenges. In turn, you'll uncover new ways of thinking about how to improve your software project's maintainability.
On the Maintainable Software Podcast, Robby speaks with seasoned practitioners who have overcome the technical and cultural problems often associated with software development.
Our guests will share stories in each episode and outline tangible, real-world approaches to software challenges. In turn, you'll uncover new ways of thinking about how to improve your software project's maintainability.
Episodes
Mentioned books
Oct 13, 2026 • 53min
Gabriela Moreira: When 800 Tests Aren’t Enough
Gabriela Moreira, software engineer and CEO of Quint, explores why 800 AI-generated tests can inspire less confidence than 200 meaningful ones. She discusses regression tests, executable specifications, and making TLA+ more approachable. The conversation covers concurrency, model-based testing, human judgment, and a Kubernetes race condition uncovered before deployment. She also champions experimentation in AI-assisted development.
Oct 6, 2026 • 54min
Matija Šošić: Abstractions Aren’t Going Anywhere
Matija Šošić, co-founder and CEO of open-source framework Wasp, explores what makes software truly maintainable. They discuss boring code, resisting premature abstractions, Wasp’s journey from a custom DSL to TypeScript, and the recurring tug-of-war between bundled and unbundled web stacks. They also examine why AI-generated code could make frameworks, conventions, and guardrails more valuable.
Sep 29, 2026 • 54min
José Valim: Designing a Language You Can Maintain for Decades
José Valim, creator of Elixir and founder of Dashbit, explores what it takes to maintain a programming language for decades. He shares the concurrency bug that led him to Erlang and the BEAM, why Elixir favors extensibility and restraint, and how Phoenix grew independently. They also discuss Tidewave, AI-assisted open source, documentation, and rethinking dependencies and SDKs.
Sep 8, 2026 • 1h 22min
John Athayde: Maintaining the User Side of Software
John Athayde, product designer and owner of Meticulous, explores maintainability beyond the codebase. They discuss documenting legacy decisions, untangling fragmented interfaces, modernizing without needless redesigns, and governing design systems. They also examine AI-assisted front-end archaeology, incremental refactoring, and the human judgment needed to preserve valuable workflows while making software easier to evolve.
Sep 1, 2026 • 54min
Kristin Isaac: The Customer Doesn’t Care About Your Technical Debt
Kristin Isaac, co-founder and CEO of Strudel, brings 20+ years of experience in customer success, sales, and support. She explores how reliability and technical debt shape customer experience, renewals, and revenue. The conversation covers prioritizing beyond the loudest customer, making invisible maintenance work matter to business leaders, and AI’s impact on software understanding, collaboration, and engineering’s role as an educator.
Aug 18, 2026 • 55min
David Hayes: Boring Software, Clear Incentives, and Better Checklists
David Hayes, a product leader and infrastructure specialist from PagerDuty, Sentry, and FusionAuth, explores why trusted software should be boring, stable, and built around customer outcomes. He discusses API design, self-hosted authentication, upgrade resistance, technical debt, scope creep, and the psychology behind estimates. Naval history, incentives, checklists, and AI-powered development add surprising perspective.
Aug 4, 2026 • 49min
Diana Pfeil: Building Confidence in Probabilistic Systems
Diana Pfeil, an ML and AI engineer at Sunbeam Consulting with experience in production ML and recommendation systems. She discusses maintaining probabilistic systems, using evals and LLMs to judge variable outputs, sampling and human review for monitoring, prompt versioning and fixes, handling sensitive data, and starting simple with prototypes before adding complexity.
Jun 16, 2026 • 55min
Chris Coyier: The Long Game of Maintaining CodePen
Chris Coyier, co-founder of CodePen and creator of CSS-Tricks, reflects on running a 15-year web product. He discusses migrating parts of the stack from Rails to Go, using GraphQL across implementations, and trade-offs around hosting costs and team size. Short takes cover technical debt, simplifying with browser features and CSS, and how small teams decide what to maintain vs rebuild.
11 snips
May 5, 2026 • 56min
Sally Lait: Confidence Is the Real Metric
Sally Lait, a fractional technology leader with 20+ years modernizing legacy systems and guiding teams. She reframes maintainability around documentation, onboarding speed, knowledge sharing, and especially team confidence. The conversation covers measuring qualitative health, communicating small wins, modernizing long-lived apps, and practical onboarding tips for older codebases.
Apr 14, 2026 • 55min
Rein Henrichs: The Real Work of Maintenance Happens Before You Touch the Code
Rein Henrichs, a principal software engineer focused on maintainability and resilience, shares why maintenance is about shared understanding. He explains the line of representation, common ground, weak signals of decaying understanding, incident coordination, partial failures, and how mentoring and observability build adaptive capacity.
