About The Agentic Company
I write for people who have to make agentic AI work
I am Erez Eden, the editor of The Agentic Company. I use this site to turn agentic AI into useful architecture, working patterns, and honest comparisons for the people building it inside real companies.
Erez Eden
I focus on the parts that decide whether an agent system is useful after the demo: context and memory, tool integration, workflow design, permissions, evaluation, cost, and governance. I also publish practical work from products I build, including the knowledge-network product Fungal.
My aim is simple: help technical founders, architects, engineers, and operations leaders make better decisions without the usual AI hype.
What I publish
The site covers enterprise context engines, memory, the Model Context Protocol, agent workflows, security and governance, evaluations, and the tools used to put those systems into production. The formats range from explainers and how-to guides to scored reviews, architecture comparisons, and evidence syntheses.
How the research works
I separate what I tested from what I learned through documentation and other primary sources. When an article includes a benchmark, working artifact, or hands-on result, I say how it was produced. When it relies on outside evidence, I link the source inline so you can check it yourself. Dates matter too: time-sensitive claims are dated, and material updates are reflected in the article's updated date.
Corrections and updates
If something is wrong, unclear, or out of date, send me the article URL and the point to check. I review the source, correct material errors, and update the article date when the change affects the conclusion or guidance. I do not quietly leave a known error in place.