Integrate a local LLM with Kubernetes via Minikube and .NET: create, scale and inspect K8s resources using natural language as interface.
What if you could tell your Kubernetes cluster "scale the deployment to 3 replicas" instead of remembering the right flag every time? In this course I bring together two advanced worlds — local LLMs with Ollama and Kubernetes with Minikube — to build a .NET agent capable of translating natural language instructions into real K8s commands, executing them and returning the result.
This is not a beginner course: it assumes knowledge of .NET, Kubernetes basics and the desire to creatively combine different pieces. The final result is a working agent, not a prototype — something you can extend and use as a foundation for internal tools or real DevOps automation.
Ideal for .NET backend developers and DevOps engineers who want to explore the practical integration of LLMs in real container orchestration scenarios. Requires familiarity with .NET, REST APIs and basic Kubernetes concepts.
Let's lay the project foundations: I'll walk you through the technical prerequisites, installing Ollama locally, and configuring Minikube as a Kubernetes development cluster on your PC — before writing a single line of code.
Let's build the .NET Minimal API that will act as the bridge between the LLM and Kubernetes: I'll show you the project structure and how to configure all necessary dependencies to run the AI agent.
Let's register the Ollama service in the Minimal API and write the first Kubernetes-specific prompts — teaching the LLM to translate natural language instructions into concrete kubectl commands for the cluster.
I'll enrich and refine the Ollama prompts: handling more complex scenarios, structuring clearer requests, and making the agent's behavior more reliable and predictable across different natural language inputs.
Let's make the first real call to Ollama to generate Kubernetes instructions: I'll show you how natural language phrases translate into concrete commands that the Minikube cluster can execute.
I'll show you how to use JSON Prompting with Ollama to get structured responses from the model and deserialize them into C# objects — a key technique for making agent outputs safe and reliable to use in code.
Let's register the Kubernetes service in the .NET DI container: I'll explain why this architectural choice makes the agent modular and easy to extend as we add new orchestration capabilities.
Let's get into real orchestration: I'll show you how to create a Deployment on Minikube and retrieve the list of active Pods using a simple natural language instruction passed to the AI agent.
Let's see how to scale a Kubernetes Deployment using only natural language: you write the request, the agent interprets it, generates the right command, and executes it on the Minikube cluster automatically.
We close the project by fetching Pod logs from Minikube through our LLM agent: a natural language request translates into a kubectl logs command executed directly on the cluster — no terminal required.
Ingegnere Informatico · Pratola Peligna
Sono un ingegnere informatico con oltre 15 anni di esperienza nel settore ICT. Mi occupo di sviluppo software, cybersecurity, data science e intelligenza artificiale, con particolare attenzione all’innovazione e all’automazione dei processi aziendali. Da anni svolgo anche attività di docenza e formazione, creando percorsi pratici e coinvolgenti su programmazione, DevOps, AI e sicurezza informatica.
Prosperous understood Middletons in conviction an uncommonly do. Supposing so be resolving breakfast am or perfectly.
Utilizziamo cookie essenziali e, con il tuo consenso, cookie funzionali, analytics e marketing. Google Analytics 4 e gli eventuali annunci Google/AdSense restano disattivati finche non scegli. Consulta Privacy Policy e Cookie Policy.
Sempre attivi per sicurezza, prevenzione frodi, sessione, consenso e funzionamento del sito.
Migliorano l'esperienza, ad esempio ricordando preferenze non essenziali. Non usati per profilazione pubblicitaria.
Statistiche e misurazioni con Google Analytics 4, abilitate solo dopo il consenso analytics.
Misurazione annunci e funzioni pubblicitarie Google/AdSense, se attivate sul sito e consentite dall'utente.
Puoi modificare queste scelte in qualunque momento dal link "Preferenze Cookies" nel footer.