Shorter field notes and technical arguments. These posts are meant to keep useful context close to the surface: architecture decisions, AI workflows, infrastructure lessons, DeFi observations, and the product details that compound into better systems.
- The “SaaSpocalypse” Isn’t Real - durable software systems still matter when LLM interfaces get better.
- ML y MLOps: notas de aprendizaje 1 - a corrected learning notebook on PyTorch, activations, gradient descent, MLOps, GPUs, stochastic inference, and agents.
- Agentic Workflows - how to think about agents as operational systems instead of demos.
- Tailscale en produccion - practical networking notes from shipping and operating systems.
- Primero hazlo naive - start with the simple implementation, then earn the abstraction.