The latest articles on DEV Community by Zakaria Hili (@zakaria_hili_686648600332). https://dev.to/zakaria_hili_686648600332 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3811062%2Fc49eee2d-6363-491f-ab29-d8468643de20.png https://dev.to/zakaria_hili_686648600332 en Zakaria Hili Sat, 07 Mar 2026 06:27:25 +0000 https://dev.to/zakaria_hili_686648600332/reducing-the-time-between-a-production-crash-and-a-fix-21pe https://dev.to/zakaria_hili_686648600332/reducing-the-time-between-a-production-crash-and-a-fix-21pe <p>You ship code, everything works — and then suddenly a crash appears in production.</p> <p>Even in well-instrumented systems, the investigation process often looks like this:</p> <ol> <li>check the monitoring alert</li> <li>dig through logs</li> <li>search the codebase</li> <li>try to reproduce the issue</li> <li>write a fix</li> <li>open a pull request</li> </ol> <p>In many teams, this process can easily take hours.</p> <p>After several years working on complex applications and critical data workflows, I started wondering if part of this investigation process could be automated.</p> <p>Could we shorten the loop between <strong>crash detection and a validated fix</strong>?</p> <p>This is what led me to start building <strong>Crashloom</strong>.</p> <p>Crashloom is an experiment around using <strong>AI agents to investigate crashes, identify potential root causes, and propose fixes that can be validated before creating a pull request</strong>.</p> <p>The idea is to reduce the time between a production crash and a safe fix by assisting developers in the investigation workflow.</p> <p>crash → investigation → sandbox validation → pull request</p> <p>The project is still early stage, and I'm curious how other teams handle production incidents today.</p> <p>How long does it usually take in your case to go from crash detection → <strong>merged fix</strong>?</p> productivity devops ai backend