From the Field

Practitioner's Log

Build logs and field notes from shipping enterprise AI — the patterns that held, the tools that scaled, and the lessons that cost the most to learn.

001

June 12, 2026

Context Engineering

Repeatable Pattern

Prompts in Code Is a Trap. Here Are Two Ways Out.

Enterprises deploying multiple agents across multiple LLM providers need prompt management and model routing as infrastructure, not afterthoughts. Tier-based model selection is the architecture that supports bring-your-own-LLM products.

002

June 12, 2026

Evaluation & Observability

Proof of Concept

I Built an LLM Observability Dashboard With a Single Prompt. Here Is What It Taught Me About What Actually Matters.

Enterprises cannot govern, optimize, or trust AI at scale without per-call visibility. System-level monitoring is not sufficient. Observability must operate at the interaction level.