Exploring Cloud Intelligence through practice and continuous Learning
I’m an IT Support and NOC professional focused on monitoring, incident response, and Microsoft cloud environments. I’m transitioning into Cloud and AI, learning through hands on practice in AWS and Azure while sharing practical insights about Cloud Intelligence, automation, and modern infrastructure. Certified AZ 900 and AI 900, continuously learning and improving through real world experience.
Markdown supported · ⌘↵ to post
Mateo Ruiz
I like that this post doesn't turn it into a local-vs-cloud debate. They're really different stages of the same journey.
Running Ollama locally is one of the fastest ways to understand what AI workloads actually cost in terms of memory, compute, and latency. A lot of those realities are hidden when you're only consuming APIs.
What I've seen in practice is that local models are great for experimentation, internal tools, and privacy-sensitive workflows, while cloud infrastructure becomes important once reliability, concurrency, and operational support matter. Most teams will probably end up with some hybrid mix rather than choosing one side forever.
The best lesson from running models locally isn't saving API costs it's learning what your future production architecture will eventually have to handle.