<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>ivomarino.com</title><link>https://ivomarino.com/</link><description>Recent content on ivomarino.com</description><generator>Hugo -- gohugo.io</generator><language>en-US</language><copyright>All rights reserved - 2026</copyright><lastBuildDate>Sun, 18 Oct 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://ivomarino.com/index.xml" rel="self" type="application/rss+xml"/><item><title>Privacy</title><link>https://ivomarino.com/privacy/</link><pubDate>Tue, 20 Oct 2026 20:00:00 +0200</pubDate><guid>https://ivomarino.com/privacy/</guid><description>Last updated: 20 October 2026.
This page covers ivomarino.com and the small personal tools I run under this domain. I am Ivo Marino, Baar, Switzerland. Questions go through the message form on the contact page.
This website The site is static. It has no accounts, no comment section and no forms that store anything. I measure visits with Umami, which does not use cookies. It records page views, referrer addresses, browser, operating system, device type and approximate location. If you arrive through a link with a short version code, for example ?cv=A2610211, Umami also records that code so I can see which version of my CV the link came from. If you write to me by email, I keep the message for as long as the conversation needs it and do not pass it on.
My Google API client I use a private Google API client (shown to Google as &amp;ldquo;amux&amp;rdquo;) so that command-line tools on my own computers can read and manage my own Gmail mailbox.</description></item><item><title>Terms</title><link>https://ivomarino.com/terms/</link><pubDate>Tue, 20 Oct 2026 20:00:00 +0200</pubDate><guid>https://ivomarino.com/terms/</guid><description>Last updated: 20 October 2026.
ivomarino.com is the personal site of Ivo Marino, Baar, Switzerland.
Content The articles and guides are my own work and reflect my experience at the time of writing. Check commands and configuration against your own environment before you use them in production. I provide them as they are, without warranty, and I am not liable for damage from using them.
Copyright The text and images on this site are copyright of Ivo Marino unless marked otherwise. You may quote short passages with a link back to the source. Anything beyond that needs my written permission.
Links The site links to external sites that I do not control. I am not responsible for their content.
Tools published under this domain The private tools I run under this domain are for my own use only. They are not a public service, and no one else is invited to sign in to them.</description></item><item><title>Build log 1: The five checks I put around an AI workflow before it touches real work</title><link>https://ivomarino.com/post/ai-workflow-build-log-1/</link><pubDate>Sun, 18 Oct 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/ai-workflow-build-log-1/</guid><description>A new series where I build AI workflows end to end and write down how. Who it is for, the problem it solves, its goals, and five checks that sit around the model.</description></item><item><title>Work with me</title><link>https://ivomarino.com/work-with-me/</link><pubDate>Sat, 17 Oct 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/work-with-me/</guid><description>Most AI projects stall between a demo and something a team can rely on. I take it from one to the other, and I stay for the part after launch.
What I build AI workflows that run a real process. An agent or pipeline that does a job end to end, for example triaging requests, drafting documents or reconciling data, with a person approving the steps that matter. MCP servers and tool access. Your internal systems exposed to AI agents with authentication, per-tool permissions and an audit log. Serving and operations. Model hosting on Kubernetes, tracing, cost per request, rollback. The same standard as any production system. Evaluation. A test suite for AI behaviour that runs on every change, so quality cannot drift without anyone noticing. Making your team the supervisors AI does the first draft; people decide. I run short working sessions that teach a team to review AI output, write good instructions, spot failure modes and know when to stop a run.</description></item><item><title>Out-of-Band SSH Tunnel Failover: Why Stability Beats Speed</title><link>https://ivomarino.com/post/gateway-failover-ssh-tunnels/</link><pubDate>Thu, 23 Apr 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/gateway-failover-ssh-tunnels/</guid><description>Building reliable automatic gateway failover for critical infrastructure recovery access. How naive health checks broke everything, and why a 30-second stable recovery is better than a 5-second flapping failover.</description></item><item><title>Kubernetes Pod Autoscaling with KEDA: Queue Depth Beats CPU Every Time</title><link>https://ivomarino.com/post/keda-queue-autoscaling/</link><pubDate>Tue, 31 Mar 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/keda-queue-autoscaling/</guid><description>Kubernetes HPA with CPU metrics fails for queue-based workloads. Learn why queue depth is the right signal, how KEDA makes event-driven autoscaling trivial, and the configuration that reduced processing latency from 45 minutes to 5-15 minutes with zero manual scaling.</description></item><item><title>Production Security Updates Without Downtime: Debian, FreeBSD, and OpenBSD</title><link>https://ivomarino.com/post/production-security-updates/</link><pubDate>Sat, 28 Mar 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/production-security-updates/</guid><description>Security patches can&amp;rsquo;t wait, but unplanned reboots break production. The update runbook for Debian, FreeBSD, and OpenBSD — with real scripts, hold patterns, and a zero-downtime cluster update sequence.</description></item><item><title>From Data Science to Production: MLOps CI/CD Pipeline with Real-time Accuracy Monitoring</title><link>https://ivomarino.com/post/mlops-guide/</link><pubDate>Fri, 27 Mar 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/mlops-guide/</guid><description>Production machine learning requires more than training good models. This guide covers building a complete MLOps system with Kubernetes, Prometheus monitoring, and automated accuracy checks to catch model drift and prevent silent failures in production.</description></item><item><title>ZFS + NFS as Kubernetes Storage: Point-in-Time Recovery Without the Cost</title><link>https://ivomarino.com/post/zfs-nfs-kubernetes-storage/</link><pubDate>Thu, 26 Mar 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/zfs-nfs-kubernetes-storage/</guid><description>Kubernetes storage is either ephemeral or expensive. We run a FreeBSD VM with ZFS + NFS in the same VPC as our K8s cluster. Hourly snapshots protect your data. Fast recovery. No vendor lock-in. Here&amp;rsquo;s how we do it.</description></item><item><title>OpenBSD CARP Firewalls in Front of a Private Kubernetes Cluster</title><link>https://ivomarino.com/post/openbsd-carp-k8s-firewall/</link><pubDate>Wed, 25 Mar 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/openbsd-carp-k8s-firewall/</guid><description>We run two OpenBSD firewalls in CARP HA mode in front of a private Kubernetes cluster. Sub-second failover, full state sync, and pf rules that load-balance across worker nodes.</description></item><item><title>Scaling MySQL Beyond Kubernetes: When Backups Become The Blocker</title><link>https://ivomarino.com/post/mysql-scaling-beyond-k8s/</link><pubDate>Mon, 23 Mar 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/mysql-scaling-beyond-k8s/</guid><description>We ran MySQL in Kubernetes for years until the data scaled from GB to TB. Kubernetes wasn&amp;rsquo;t the problem—backups were. Kasten (the Kubernetes backup tool) was never designed for databases at our scale. Learn why we moved MySQL off Kubernetes and the backup solution we built to solve this infrastructure challenge.</description></item><item><title>2026 Relaunch: Infrastructure Lessons from 10+ Years Running Production</title><link>https://ivomarino.com/post/relaunch/</link><pubDate>Thu, 01 Jan 2026 00:00:00 +0000</pubDate><guid>https://ivomarino.com/post/relaunch/</guid><description>Relaunching with a focus on pragmatic infrastructure: backup strategies, Kubernetes autoscaling, and lessons from 10+ years of production operations.</description></item><item><title>Contact</title><link>https://ivomarino.com/contact/</link><pubDate>Sat, 13 Apr 2019 12:12:20 -0700</pubDate><guid>https://ivomarino.com/contact/</guid><description>Book a free 30-minute video call, or send me a message if you prefer to write.
Book a call or send a message Book a call Send a message Day Time
Video Name Email What would you like to talk about? (optional) How did you find me? (optional) Website I agree that my name and email are used to schedule this call. Name, email and note are deleted 90 days after the call.
Book the call Loading free times…
Booked The calendar invitation is on its way to your inbox.
WhenWhere Join link
No call needed? Tell me what you are working on and I will answer by email within one working day.
Name Email What is it about? Platform, Kubernetes or DevOps Cloud migration An app or service A website Something else Budget (optional) Message How did you find me? (optional) Website I agree that my name, email and message are used to answer me.</description></item><item><title>About me</title><link>https://ivomarino.com/about/</link><pubDate>Sat, 13 Apr 2019 11:06:24 -0700</pubDate><guid>https://ivomarino.com/about/</guid><description>Born in Germany, raised and studied in Italy. After graduating from the University of Rome with a BS in Computer Science, I spent 15+ years building and operating production platforms for Swiss enterprises and startups.
Specialised in Kubernetes platforms across multi-cloud environments (AWS, Azure, GCP). Focused on infrastructure automation (Terraform, Ansible), GitOps practices (ArgoCD), and operational excellence. Founder of floads, providing platform engineering and infrastructure consulting to Swiss enterprises.
I work best with passionate, pragmatic teams solving real infrastructure challenges—not theoretical problems. Strong believer in learning by doing, infrastructure as code, and automating away tedious operations.</description></item><item><title>Resume</title><link>https://ivomarino.com/resume/</link><pubDate>Sat, 13 Apr 2019 00:00:00 +0000</pubDate><guid>https://ivomarino.com/resume/</guid><description>Senior Platform Engineer / SRE with 15+ years building and operating production platforms, specializing in Kubernetes across multi-cloud environments for enterprise and SaaS workloads. Currently building the GPU inference infrastructure behind a production AI platform. Specialised in Kubernetes (EKS, K3s, multi-cluster), GitOps (ArgoCD), and Infrastructure as Code (Terraform, Ansible) across AWS, Azure, and GCP.
Delivered production-grade platforms for Swiss enterprise customers including Allianz Technology, NTT, REWE, Flow Swiss, and LerNetz+ AG. Focused on reliability, scalability, cost-efficient platform design, and improving developer experience.
Fluent in German (native), Italian (native), and English. Founder of floads — providing Kubernetes and infrastructure consulting to Swiss enterprises.
Work Experience Senior Infrastructure Engineer — AI Platform (contract) 2026 - present Building the GPU inference infrastructure behind a production AI video platform. Migrating GPU-backed model serving (NVIDIA Triton) from AWS ECS to multi-region EKS, including Karpenter GPU node pools on Bottlerocket and rollout strategies for single- and multi-model inference services.</description></item></channel></rss>