abeebaNotes from the infrastructure floor.
Practical articles on private AI, GPU economics, MLOps and AI security, written by the engineers who do the work.
Private LLM or public API? A simple cost model
When does running your own model become cheaper than paying per token? A practical way to work it out with your own numbers.
7 min readFive ways to cut inference cost without changing your model
Before you swap models or renegotiate contracts, these engineering changes usually deliver the fastest savings.
6 min readAir-gapped LLM deployments: a field checklist
In progress. Subscribe below to get it when it is published.
Choosing between LoRA and full fine-tuning
In progress. Subscribe below to get it when it is published.
What auditors ask about your AI systems
In progress. Subscribe below to get it when it is published.
Kubernetes GPU scheduling for shared clusters
In progress. Subscribe below to get it when it is published.
Get new articles by email.
One email when something new is published. No spam, unsubscribe any time.
Ready to run AI on your own terms?
Book a 30-minute consultation with an infrastructure engineer. You'll leave with a clear view of options, costs and next steps, whether or not you work with us.