Become the engineer who can take AI
from demo to production
Build and operate a production-shaped AI platform on Kubernetes in four weeks — model serving, compute, observability, security, automation and cost. For experienced DevOps, cloud, SRE and platform engineers.
Four weeks · Live instruction · Practical labs · Portfolio capstone · Limited founding cohort
AI changes the workload. It does not remove the need for engineering.
Many teams can connect an application to a model API. Far fewer can answer: how should the workload be deployed, how will it scale, how will access be controlled, how will failures be detected, how will GPU and inference costs be measured, and what makes it ready for production?
These responsibilities increasingly reach DevOps, cloud and platform teams. This bootcamp helps you extend the skills you already have into this new operating environment.
By the end, you will be able to:
Explain the architecture of a production AI application
Deploy and operate model-serving workloads
Use Kubernetes to manage AI infrastructure
Understand GPU scheduling and capacity decisions
Implement meaningful observability
Evaluate security and production-readiness risks
Estimate operating cost
Present an AI-platform recommendation to technical leadership
You will build — not merely watch
Your capstone includes:
A functioning model-serving endpoint
Kubernetes deployment configuration
Automated infrastructure setup
An observability dashboard
A security and readiness checklist
A cost model
An operational runbook
An architecture presentation
Use the finished project to demonstrate your capability internally, during interviews, or with clients.
Four weeks from infrastructure experience to AI-platform capability
Model to service
Understand inference architecture — model APIs, gateways, latency and availability — and deploy the first working model-serving workload.
Deliverable: Architecture diagram + functioning inference endpoint
Kubernetes and compute
Schedule AI workloads, size CPU/GPU resources, share and isolate capacity, distribute models via storage, and automate the environment.
Deliverable: Kubernetes deployment + automated environment configuration
Production readiness
Add authentication, secrets and supply-chain controls, AI-specific observability, failure handling, rollback and reliability objectives.
Deliverable: Observability dashboard + operational runbook
Cost, governance and capstone
Estimate unit economics and capacity, weigh multi-tenant and governance trade-offs, and present your complete platform.
Deliverable: Complete AI-platform blueprint, cost model and presentation
An optional preparation module reviews containers, Kubernetes basics, and lab setup before Week 1 for anyone who wants a refresher.
This is for you when:
- You work in DevOps, SRE, cloud or platform engineering
- You understand infrastructure fundamentals
- AI workloads are beginning to reach your team
- You want practical evidence of your capability
- You learn best by building
- You want to remain valuable as infrastructure work evolves
This is not designed for:
- Complete technology beginners
- People seeking only prompt-engineering techniques
- Data scientists looking for modelling instruction
- People wanting a passive video course
- Anyone expecting a guaranteed job outcome
Learn with Luca Berton
Luca is an AI Platform Engineering Educator, KubeCon speaker, Docker Captain, former Red Hat engineer and author of eight technical books. He has taught 40,000+ students across automation, Kubernetes, cloud infrastructure and AI-assisted engineering.
His teaching focuses on real systems, practical implementation, and the operational details that are often missing from high-level AI content.
Format
- Four-week live cohort
- Two 90-minute sessions per week
- Weekly office hour
- Practical labs
- Private participant community
- Session recordings
- Reusable templates and repository
- Capstone review
- Completion certificate
Time commitment
Approximately five hours per week: three hours live, two hours lab and capstone work. The capstone grows incrementally week over week, so you're not starting a large final project at the end.
Prerequisites
Linux command-line familiarity, basic Git experience, container fundamentals, and basic Kubernetes awareness. An optional preparation module reviews the concepts used in the labs.
Founding-cohort pricing
Applications open now — founding cohort date confirmed by email once seats fill. Founding pricing is tied to your feedback helping refine the programme.
Plus
€895
- Everything in Standard
- Individual capstone review
- 30-minute career or architecture consultation
- Personalised skills-gap recommendation
Attend the first two live sessions and complete the first lab. If the programme clearly doesn't match the published level or curriculum, request a refund before the third session.
Frequently asked questions
Do I need a GPU?
Do I need to be a data scientist?
How much Kubernetes experience do I need?
Will sessions be recorded?
Can my employer pay?
Does the programme guarantee a job?
Which tools will we use?
When does the founding cohort start?
Apply for the Founding Cohort
A few minutes to apply. It doesn't commit you to purchasing, and Luca reads every application personally.
The next generation of AI systems still needs infrastructure engineers
The question is whether you'll be ready to operate them.