Miguel Ausó — Engineering Notes
Learning by building. Scaling through writing.
This site collects what I learn while building and operating real systems — Docker security, network isolation, LLM fine-tuning, and running AI agents in production. The documentation sections are practical playbooks for specific setups I run. The blog covers decisions, tradeoffs, and things I would rather not have to rediscover later.
Documentation
Mistral Fine-Tuning Lab
A practical walkthrough of the Mistral fine-tuning pipeline I used, from dataset prep and ChatML tokenization to QLoRA training and inference testing.
OpenClaw Operations
OpenClaw operations documentation for secure deployment, network isolation, cost optimization, and day-2 operational practices.
Recent Posts
View all →OpenClaw Docs
I finally turned my OpenClaw notes into actual documentation: the deployment setup, the network isolation choices, and the operational bits that were too easy to forget.
Mistral Fine-Tuning Lab, Documented End to End
A new documentation set covering dataset preparation, ChatML tokenization, QLoRA training, and inference for a Mistral fine-tuning workflow.
7 Mistakes That Doom AI Agents in Production
The most common failure modes for autonomous AI agents and how to prevent them before they cost you.