When an AI Agent Chooses Where a Microscope Looks Next
Argonne’s microscope demonstration connects a scientific goal, live images and the next measurement. Here is what changed, what already existed, and where human expertise still matters.
Researching. Building. Explaining.
Then I share what works, what breaks, and what is worth your time—across AI, automation, and developer tools.
Featured essay
Argonne’s microscope demonstration connects a scientific goal, live images and the next measurement. Here is what changed, what already existed, and where human expertise still matters.
Browse by interest
Models, orchestration, voice workflows, and useful experiments.
02Workflows that reduce friction without hiding the trade-offs.
03Self-hosting, infrastructure, privacy, and maintainable tools.
04Implementation lessons, frameworks, and practical code.
Recent writing
I moved my reusable AI workflows into GitHub so the skill belongs to me rather than to one AI platform. Here’s why portability, versioning and a single source of truth matter.
I built my blogging agent by correcting one frustrating behaviour at a time. This is the practical workflow that turns recurring discussions into a reviewed article, approved visuals and a verified WordPress draft.
Our everyday questions, experiments and unfinished thoughts may already contain the stories we should be telling. Category: Practical AI Systems | Estimated reading time: 6 minutes For a…
Genkit and LangGraph overlap, but they solve different layers of an AI application. The better choice depends on whether the difficult problem is building the AI capability or controlling a stateful process over time.
For a long time I was searching for a good notetaking app. Handy, something like Notion, I loved the databases, Views and Writing Assistance features in it. But…
I used to wonder about people who respond instantly. You ask them something — they reply immediately with clear statements, without fluff. No awkward pause. No “umm.” No…
About the writer
I am a software engineer, technical leader, and persistent explorer of practical AI. This is where I document what I build, what breaks, and what I learn.