Blog
Thoughts on cloud, AI and DevOps
Practical write-ups for small teams, in plain English. A new post every week.
Browse by topic
Start here: hands-on guides
Hands-on guide · AWS · US infrastructure
How to Deploy Llama 3.1 on AWS EC2 (g4dn.xlarge) for Under $1/Hour
A complete walkthrough with measured results: 34 tokens per second at about $0.53 an hour.
Hands-on guide · UpCloud · European infrastructure
Deploying Llama 3.1 on UpCloud (NVIDIA L4) with llama.cpp
The same model on EU-hosted hardware: 50.5 tokens per second, for clients who can’t use US infrastructure.
All posts
The Infrastructure Mistakes Startups Make in Year One on AWS
A founder with AWS startup credits and a product ready for launch is in a good position. They have twelve months of…
The DevOps Tools Everyone Is Hiring For, and What a Small Team Actually Needs
Job postings for DevOps roles are starting to look the same. They all list Kubernetes, Terraform, Ansible, Grafana, Prometheus, and Helm. Sometimes…
The CI/CD Pipeline a Small Software Team Actually Needs
A version of CI/CD is constantly written about. It includes automated testing across multiple environments, canary deployments, feature flags, rollback pipelines, and…
When It’s Time to Stop Hosting It Yourself
Self-hosting has many advantages. You have full control, costs are predictable, and you are not dependent on a platform you don’t own.…
Your First Cloud Setup: What Actually Needs to Be Decided Before You Touch a Provider
Most first-time cloud setups go wrong in the same way. Someone opens an AWS or DigitalOcean account, starts clicking around, and two…
What I Ask Before Setting Up Any AI Workflow for a Client
Every AI project I’ve worked on has started the same way. A client would come to us with an idea for a…
How to Actually Evaluate an AI Tool Before You Pay for It
In the last eighteen months, I have had more conversations about AI tools than about any other topic. Most of these conversations…
Local AI vs. the Cloud: The Moment That’s Actually Here
For the past two years, the conversation about AI has mostly been about access. Which API? Which model tier? Which provider? The…
The WordPress Redesign Trend That Most of My Clients Should Skip
Last month, a client sent me a link. It was a competitor’s redesigned site with huge, oversized headlines, text that animates on…
