About Alchemic Technology
Updated 6 November 2026.
Alchemic Technology designs and operates AI agent systems, on your infrastructure or ours. Manual work in. Automated gold out.
Who we are
Alchemic Technology designs, builds and operates AI agent systems for businesses that have repetitive manual work and want it done reliably. That means chatbots, workflow automation, custom integrations and multi-agent orchestration.
An engagement can run on infrastructure the client controls or on infrastructure we operate. The service is designed so a person approves an agent draft before it is sent, and so a later decision to let an agent act alone is written down. This page describes that design.
The guides and the blog focus on the same areas as that design: what an agent is allowed to do, how a team evaluates it, and what should happen when a run is no longer cleared to act. A reading guide for the safety posts is at Agent safety and permissions.
What we build
We build four kinds of system. Chatbots answer from your own material and hand off to a person when a question needs one. Workflow automation connects tools such as Slack, GitHub, Linear and a CRM. Custom integrations move data between systems. Orchestration gives each of several agents a defined job, defined permissions and a point where a person steps in.
One example we use to describe the work, the same kind of example as on the homepage: an agent reads an inbound support email and drafts the reply, and a person approves it before it sends. The published engagement path uses four stage names from the alchemist's process: Nigredo, Albedo, Citrinitas and Rubedo.
We also publish practical guides, including the Agent Readiness Kit: 14 editable templates for deciding what an AI agent may do on its own.
How the blog is written
The blog covers agent safety, evaluation, governance and the cost of running AI yourself. A post is a synthesis or an opinion unless it says it is summarizing a named document.
Where a post names a paper, specification, documentation page or pricing page, that post should link the document so you can check the sentence. Posts revised on 6 November 2026 do that for the documents they name. Those posts are the quarantine essay and the six posts collected in the agent safety and permissions reading guide. Many earlier posts still name a source and leave it unlinked.
We use AI tools to help research and draft posts and page copy. Alchemic Technology is responsible for what it publishes on this site.
If you find an error, email jjmrowic@alchemictechnology.com. We will correct the post and note the change.
Get in touch
Questions, corrections or project inquiries: jjmrowic@alchemictechnology.com or use the contact page.