Supporting developers working on a large and complex C++ application isn't easy. Long build and iteration times, cryptic compiler errors, platform-specific dependencies, and complex merge issues can all turn a seemingly small code change into a lengthy feedback loop. At Autodesk Fusion, we've been working on reducing that friction using a combination of traditional engineering and AI. In this talk, I'll share a few practical examples: how we built a custom caching solution for C++ build artifacts that can support AI workflows, how we're using AI to avoid unnecessary build work, and how AI can analyze build logs, explain failures, and help act on the issues it finds. Rather than treating AI as a replacement for build infrastructure, I'll show how we're combining it with build-system knowledge and automation to shorten the feedback loop for developers working in a large C++ codebase.
Piotr is a Senior Principal Engineer at Autodesk, focused on supporting everything that can be built locally or in CI within the Autodesk Fusion ecosystem.
Piotr has over 20 years of experience in software development. He started his career as a C++ systems programmer in game development and got “tricked” into CI/CD and build pipelines in 2010. Since then, he has worked in a variety of environments, ranging from fast-paced, experimental game development to “serious” enterprise software at Oracle MySQL and Autodesk Fusion.
