Prakash prescribes failure‑first engineering checklist

- Prakash Sharma advised engineers to list failure scenarios first — bad inputs, scaling limits, and integration failures — before coding starts and remediation plans. - He says designing safeguards and rollback paths early reduces costly rework and keeps teams focused on observable failure modes and testing. - The failure‑first stance pairs with chaos testing and SLI-driven design to improve release confidence operationally. (x.com)