A Friday maintenance window that almost turned into breach response A platform team was rolling out a minor patch to internal Windows jump hosts. The change itself was safe. The risk came from the shortcut around it: an engineer enabled…
Author: Ankur Sharma
-
The Script That Broke in CI: A 2026 Python Playbook for PEP 723, uv Script Lockfiles, and Reproducible One-File Automation
PEP 723 and uv script lockfiles make one-file Python automation reproducible across laptops and CI. Learn a practical migration path, tradeoffs, and fixes.
-
The Instant Search That Froze on Mid-Range Phones: Frontend Performance Engineering with CPU Budgets and Main-Thread Backpressure
A launch-day bug that never showed up on developer machines A B2B SaaS team shipped a new “instant search” experience in their web app. On fast laptops it felt fantastic, nearly native. In production, support tickets started within hours: “typing…
-
The Restore Drill That Exposed Empty Backups: Building Immutable Cloud Backups with S3 Object Lock and AWS Backup Vault Lock
Build immutable cloud backups with S3 Object Lock, AWS Backup Vault Lock, and restore testing so incidents turn into controlled recovery, not data-loss chaos.
-
The Partial Commit Gap: A 2026 Backend Reliability Blueprint with Outbox, Inbox, and Deterministic Replay
A small shipping delay that exposed a big reliability hole A logistics startup had a classic “everything looks green” morning. API uptime was fine, queue throughput was normal, and database CPU was low. But customer support tickets kept coming in:…
-
The 2 AM AssumeRole Failure: A Multi-Account GitHub Actions OIDC Runbook with Session Policies and Break-Glass Controls
GitHub Actions OIDC multi-account AWS runbook with strict IAM trust policies, session controls, and break-glass safeguards for auditable, safer deployments.
-
The Memory Layer That Changed the Answer: An AI/ML Production Playbook for Reproducible Agent Behavior in 2026
A production bug that looked like model randomness A support automation team rolled out an agent that drafted replies, linked policy docs, and escalated risky requests. It worked well in staging. In production, two agents answered the same customer question…
-
The Thread Pool That Stopped Scaling: A Java 21 Virtual Threads Migration Runbook for Spring Boot APIs
A practical Java 21 virtual threads migration runbook for Spring Boot APIs, covering pinning, database pool tradeoffs, and safe backpressure in production.
-
The Knowledge Drift Problem in WordPress: A 2026 Engineering Playbook for Git-Native Content Ops and Safer Automation
A small publishing error that became a big trust issue A media site migrated to AI-assisted editorial workflows to speed up publishing. It worked at first. Drafts were faster, summaries looked polished, and editors loved the reduced manual work. Then…
-
The Dashboard That Changed After Lunch: An Iceberg Snapshot Audit Workflow with Spark, DuckDB, and dbt
Practical Iceberg snapshot audit workflow using Spark, DuckDB, and dbt to trace metric drift, validate backfills, and debug data changes with confidence.
-
The Backfill That Changed Revenue History: SQL Data Engineering Patterns for Safe Reprocessing in 2026
A Monday morning surprise from a “routine” backfill A growth team asked for a simple fix: “Can we backfill missing purchase events from last month?” The data engineering team ran the job, dashboards refreshed, and everyone moved on. Two hours…
-
The DOM XSS Backlog Trap: A Website Security Playbook for Migrating to Trusted Types with Measurable Risk Reduction
A practical Trusted Types migration and strict CSP rollout plan to reduce DOM XSS risk, move from report-only to enforcement, and avoid production breakage.