short bio
steve preu is an agentic systems architect in austria. steve designs and builds agent harnesses, memory and deterministic workflows, so ai agents and the people working with them start each session from what the last one learned. steve wrote pm cli, open-source, git-native project management for humans and coding agents, and publishes it with its extensions as unbrained. open to engineering roles, consulting projects and collaborations.
third person, for event pages and introductions.
at a glance
- also known as
- stefan preu legal name
- unbraind github and tokscale
- unbrained npm and this site
- @stevepreu x
- role
- agentic systems architect
- based in
- austria, central european time
- services
- agent harnesses, memory and context, deterministic workflows, ready for agentic work · what i build
- open to
- engineering roles, consulting projects and collaborations.
- flagship
- pm cli: git-native project management for humans and coding agents
- open source
- 37 projects1,355 versions of 23 packages on npm since 4 mar 2026; releases on 159 of 231 days.
- github
- github.com/unbraind38,177 contributions on github since 20 oct 2025; active on 356 of 366 days.
- contact
- preu.at/contact, the short form on steve's personal site.
questions
short answers
who is steve preu?
an agentic systems architect in austria who designs and builds agent harnesses, memory and deterministic workflows. the legal name is stefan preu; the open-source work is published as unbrained on npm and unbraind on github.
what does an agentic systems architect do?
designs the system an ai agent works inside, rather than the model: the harness that sets its tools and permissions, the memory that carries context from one session to the next, and the workflows that fix the steps which must repeat while leaving room where an agent should investigate. the aim is agent work that a team can inspect, repeat and own.
can i hire steve preu?
yes. steve is open to engineering roles, consulting projects and collaborations, from a first build to ongoing adaptation, preferably built together with your team. the quickest way in is the short form at preu.at/contact.
what is pm cli?
steve's open-source project management for humans and coding agents. goals, acceptance criteria, decisions and history live as plain files next to the code, with every change in an append-only history, so a fresh agent session can read what happened and carry on. mit licensed, with a cli, an sdk and an mcp server.
what has steve published?
pm cli, its extensions and presets, its hosted sites pm-web, pm-gpt, pm-mcp and pm-search, and agent-ready clis for jellyfin, home assistant, monica and searxng. 37 open-source projects in all, 1,355 versions of 23 packages on npm since 4 mar 2026.
is unbrained a company?
no. unbrained is the name steve publishes open-source work under, on npm and on this site. on github it is unbraind: the same word without the e.
contact
hiring, or building something?
either way, i would like to hear about it. tell me what you are working on and where agents could help.
open to engineering roles, consulting projects and collaborations.
send a messagepreu.at/