"Build an AI assistant that answers well" — the most popular and the most useless requirement. In th

In projects with high software quality and safety requirements, engineering teams often face three major challenges: meeting project deadlines, producing the required artefacts and ensuring genuine system reliability. Various types of failure analysis are used to improve system reliability and support this objective. However, in practice, failure analyses often turn into a collection of "dead" artefacts that do little to make the software more reliable or predictable. Instead, they tend to make development processes more burdensome, creating a stream of low-value activities that reduce development speed and employee motivation.

In this talk, I will share my experience in implementing and applying failure analysis in software development and show how the process is structured. Let's discuss how the Docs-as-Code approach makes it possible to keep the analysis results up-to-date and traceable in the face of constant system changes without significantly increasing effort.