Context Engineering Evals Remind Me of Infrastructure Testing
A comparison between infrastructure testing and agent evaluation, focused on hidden state, behavioral contracts, and feedback loops.
Blogs
All public writing from Ranjib Dey on production engineering, context engineering, open source, field systems, and learning.
Planning Overlanding Trips Like Reliability Reviews
A field note on treating overlanding plans as operational reviews with explicit assumptions, risk budgets, and verification timing.
My Personal LLM Wiki Operating Model
How a personal LLM wiki can separate raw evidence, curated memory, context packs, privacy boundaries, and evaluation loops.
What Incident Taxonomies Are Really For
A reliability note on why incident categories matter only when they change prioritization, learning, and operational follow-through.
reef-pi as a Small Open Source Platform
A close look at reef-pi's hardware abstraction layer as the line that turned a controller project into a small open-source platform.
The Through-Line From Chef to Agentic AI
A systems-through-line from Chef, infrastructure automation, and TDD in operations to context engineering for agentic AI.
Treating Context Like Code
Why reusable prompts and context packs deserve versioning, review, privacy boundaries, failure modes, and evaluation loops.
reef-pi and Public Engineering Lessons
Lessons from reef-pi as a public engineering artifact that must survive real users, hardware variation, and community support.
Platform Engineering Through Automation
How automation, testing, and platform thinking connect infrastructure-as-code work with later production engineering practice.
About this website
What this website collects, how public material is selected, and where to find broader profile and project links.