The Production AI Expert
Amsterdam → worldwide
The AI demo is easy.
Production is hard.
Luca builds AI factories for enterprises, trains the engineers who run them, and brings back what works from conference stages around the world.
Planning an event? Book Luca to speak
For developers
Tested hooks, settings presets, a doctor script and a troubleshooting playbook. Install in one command.
For platform teams
Two keynote decks with speaker notes, a 90-minute lab and working agent code with guardrails.
For engineering leaders
Fixed fee, 2–6 weeks, delivered by Open Empower: the infrastructure, data, governance and compliance gaps between your AI prototype and production.
15+ years in enterprise infrastructure · 8 technical books · 40K+ students · KubeCon and Red Hat Summit speaker · Docker Captain
Where the expertise comes from
Luca sees production AI from three sides: the conferences where it is debated, the people who build it, and the client platforms where it has to run.
01 · On stage
KubeCon + CloudNativeCon, Red Hat Summit, Nordic APIs Platform Summit, Techorama, SREday: Luca speaks where production AI is being worked out, and comes home with what the best teams are doing this quarter, not last year.
02 · In conversation
On Udienza, Luca sits down with the founders, maintainers and engineers behind the platforms enterprises depend on. Recorded at conferences from Prague to San Jose, the interviews give him, and his clients, a direct line to the people shaping the stack.
The AI factory, layer by layer
03 · In production
An AI factory is the platform that lets a company ship AI again and again, not one demo at a time. Luca designs and delivers it end to end, from the GPUs to the evidence your risk team asks for.
What Luca can do for you
Three ways Luca creates value when he joins your organization, each one useful on its own.
For your engineers
Hands-on programmes that turn experienced DevOps, cloud and platform engineers into the people who can run AI in production. Real clusters, real agents, real failure modes.
For your platform
From an empty cloud account or empty racks to a governed platform your teams ship on: GPU compute, Kubernetes and OpenShift AI, model serving, agents, observability and cost control.
For your stakeholders
Luca starts where the board, security and platform teams can all see progress: a fixed-fee assessment that shows the gaps between your prototype and production, what to fix first and what it costs.
For events and sponsors
Luca brings 11,000+ LinkedIn followers, a YouTube channel and the Udienza interviews to your event: more registrations before it, more visibility for sponsors during and after.
Fill the room
LinkedIn posts and short videos for a technical audience, with a tracked link or discount code so registrations can be counted.
Speaker previews
Five-minute conversations with speakers before the event, each ending with an invitation to their session.
Sponsor interviews on site
Short videos at the booth showing a real integration or demo, clearly marked as sponsored and ready to reuse.
Coverage that lasts
Daily updates from the floor, clips for the weeks after and a recap on Udienza or this site.
The Production AI Framework
A model can answer a prompt in minutes. An enterprise still has to make that system reliable, secure, observable, affordable and governable for years. This is the operating model Luca uses to close that gap.
Prove the use case and the economics before the demo becomes architecture by accident.
Standardize model access, compute, data paths, security and developer workflows on a platform teams can operate.
Engineer for reliability, observability, performance, failure recovery and cost — not just a successful inference call.
Make governance visible with controls, policies, logs, runbooks and evidence an audit or risk team can verify.
Production AI newsletter
Practical notes on AI platforms, agents, GPUs, MLOps, governance and the operating decisions behind reliable enterprise AI.
Practical guides on Ansible, Kubernetes, Cloud Automation, and Hybrid IT Infrastructure.
by Luca Berton
Build real-world software with Claude Code — a hands-on, project-based companion covering planning, testing, refactoring, and shipping with AI.
by Luca Berton
Designing, Deploying and Scaling AI Solutions with Red Hat Enterprise Linux.
by Luca Berton & Grzegorz Stencel
Practical Guide for Container Orchestration and Deployment.
by Luca Berton
A technical guide for building secure production systems using RHEL 9 administration.
by Luca Berton
Your complete guide to Linux administration, automation, and certification.
by Luca Berton
Deep dive into advanced Ansible techniques, covering complex playbooks, modules, and custom integrations.
Upskill with professional video courses in AI, Kubernetes, Ansible, Cloud, and DevOps.
Comprehensive training for DevOps professionals in IaC, automation, and container orchestration.
Enroll Now on Udemy
Enhance your IT expertise with courses on RCA, MLOps, API security, and advanced troubleshooting.
Enroll Now on Coursera
Deep dive into Kubernetes, AI, and Ansible for advanced automation and infrastructure management.
Enroll Now on Pluralsight
Master Ansible for efficient IT automation and infrastructure management with hands-on lessons.
Enroll Now on Educative
Automate IT tasks, deploy apps, secure systems, and streamline workflows in 40 minutes.
Enroll Now on SkillshareSee what professionals and students are saying about my courses and books.
I really enjoyed this course and engaging in optional tasks...
Sharron D.
Coursera - July 31, 2024
Insightful course.
Velly H.
Coursera - December 14, 2024
Love the short clips and frequent pop questions...
Nephthalie
Coursera - August 14, 2024
Really great introduction to be honest.
Lance R.
Udemy - 4 weeks ago
If you work with Kubernetes - This is a MUST.
Erez Kirson
Amazon - June 8, 2023
Pleased to have earned a certificate for "API Validation with Postman" by Luca Berton offered through Coursera...
Nadeem Choonee 3rd+ B(Eng) | MBA | MIET | PMP | PBA | Radio/C...
Coursera - August 7, 2026
Insights on AI, Kubernetes, Ansible, and cloud infrastructure.
Build 10 real-world projects with Claude Code — Skills, hooks, MCP, testing, Git, and production workflows. My new Packt book launches December 9, 2026.
Hermes cron job failed: 'the AI model service did not respond in time. No backup provider is configured'? Add a fallback chain, unpin the job, or pause it.
Hermes Agent with a local Ollama model says 'Provider has been unresponsive' after 5 stale attempts? The model is too slow, queued or wedged. Here's the fix.
Your agent's MCP server returns HTTP 403 with a Cloudflare Access page, and 'retrying over SSE' fails too? It's Access, not the transport. Use a service token.
Etiënne vantKruys of trgc.io on AICA, a conference on AI and chips in Amsterdam that puts hardware and software builders, industry and academia in one room.
Typed a ! command in the Claude app or on claude.ai/code and got 'Shell commands are only available in local sessions'? Shell mode needs a terminal. What works.
Fresh from YouTube, TikTok, LinkedIn & Instagram.
30%+ cost reduction. 2-3× faster deployment. Compliance from day one.
Need expert guidance?
Free 30-min consultation with Luca Berton
Free 30-min Production AI consultation
Book Now