In Q3 2024, our 12-person platform team slashed log ingestion spend by 35% in 90 days, moving from a brittle Elasticsearch-based pipeline to a tuned Vector 0.30 and Loki 3.0 stack—without losing a single log or breaking our 99.95% SLA. GameStop makes $55.5B takeover offer for eBay (279 points) Talking to 35 Strangers at the Gym (144 points) Newton's law of gravity passes its biggest test (15
We Cut Compliance Costs by 40% Using Pulumi 3.140 and Chef 18 for Multi-Cloud AWS and GCP Modern multi-cloud environments offer unmatched flexibility, but they also introduce complex compliance challenges. For our team managing hybrid infrastructure across AWS and GCP, manual policy enforcement and fragmented tooling were driving up compliance costs by 22% year-over-year. By integrating Pulumi 3
In Q3 2024, our 12-person platform engineering team reduced confirmed security incidents by 41.7% (from 72 to 42 per quarter) after rolling out Trivy 0.50 for pre-deployment scanning and Falco 0.40 for runtime detection across 142 production microservices. We didn’t rewrite our CI/CD pipeline, we didn’t hire a dedicated security team, and we didn’t spend a dime on enterprise security tools. Here’s
Book: TypeScript in Production Also by me: The TypeScript Library — the 5-book collection My project: Hermes IDE | GitHub — an IDE for developers who ship with Claude Code and other AI coding tools Me: xgabriel.com | GitHub You have seen the shape of this incident before. A 500 lands in production. The frontend says "checkout failed". The Hono service that owns /checkout called the prici
Every observability vendor has bolted "AI" to their landing page. Half of those features are genuine improvements. The other half are autocomplete in a costume. After a few years of running these tools across enterprise estates, here is where AI-augmented SRE actually pays off, where it doesn't, and what we'd advise teams adopting it today. The single most defensible use case. A medium-sized estat
Iris v0.4.0 ships today. It's the release where protocol-native eval crosses from "deterministic rules" into "semantic scoring" — without giving up any of what made the deterministic layer work. Three headline features plus a lot of infrastructure work that quietly compounds. I'll go through each, why it matters, and how it fits the thesis. Heuristic rules catch a lot: length, keyword overlap, PII