Maintainable is a software engineering interview show about the work that begins after a system has users, history, deadlines, and baggage. Robbie Russell talks with seasoned practitioners about keeping software understandable and useful when easy answers have run out. The recurring focus is maintenance, but the show treats that word broadly. Code matters. So do team memory, communication, feedback loops, product pressure, management skill, and the ability to revisit old decisions without contempt. Recent conversations range from technical debt and legacy systems to AI coding agents, optionality, boring technology, database work, and the cost of running teams at full utilization. The guests tend to resist tidy slogans. Rein Henrichs frames maintenance around whether software makes sense to the people responsible for it. Russ Olsen pushes on the hidden assumptions that stick to old systems. Joel Oliveira talks about predictability, incremental migration, and compassion for past decisions. Lucas Roesler emphasizes readability and fast feedback in systems shaped by client work and complex database projects. Brittany Ellich considers AI as a tool for structured maintenance rather than wholesale replacement. Kent Beck brings the discussion back to options, capacity, and the value of making future change cheaper. Chris Zetter favors technology choices that reduce future traps. Nathan Ladd connects standards, validation, and cross-functional barriers to the daily health of a codebase. Across episodes, Maintainable feels less like a lecture series and more like a careful engineering debrief. The pacing allows nuance. The host asks practical follow-ups, especially when a concept needs to be grounded in team reality. The show is strongest when it sits with tension: measuring progress without flattening weak signals, modernizing without erasing history, and improving software without pretending every mess was avoidable. It is for engineers and technical leaders who have inherited systems, maintained systems, or wondered why a clean rewrite so rarely solves the deeper problem.