brauer.dev

Bridging AI
and Government

I work at the intersection where technology meets institutional reality — building AI capability that functions in the environments where it actually needs to work. Former Bundeswehr officer. ML engineer. GovTech strategist.

View LinkedIn Profile
01

The Question I Work On

How do you transform institutions that weren't designed to be transformed? That's the question that has followed me from NATO planning rooms to startup accelerators to banking infrastructure — and it's the question I still don't have a complete answer to.

Seven years as a Bundeswehr officer taught me how institutions actually make decisions under pressure — not how they claim to, but how they really do. That knowledge is more valuable than any technical skill I've acquired since.

Today I lead AI strategy and innovation at S-Communication Services GmbH, the central technology provider of the Sparkassen-Finanzgruppe. Before that: two AI startups, Germany's Cyber Innovation Hub, and a software deployment that took two years to navigate through Bundeswehr procurement — and is still running today.

AI Strategy GovTech Public Sector Defense ML Engineering Regulated Tech
—
Currently
Head of Innovation Management
S-Communication Services GmbH, Berlin
Teaching
Lecturer, Science, Technology and Society Studies
CODE University of Applied Sciences, Berlin
Studying
Advanced Graduate Studies in AI & ML
University of Texas at Austin
Background
Bundeswehr Officer · 7 years
NATO Operations · Intelligence · Chief of Staff
Based in
Berlin, Germany
02

What I'm Reading

001 Scott · 1998
Seeing Like a State
James C. Scott
The most useful book I've read for understanding why large-scale transformation projects fail — in government and in corporations. Scott's concept of "legibility" explains something I observed throughout my Bundeswehr years: institutions simplify reality into what they can measure and control, and that simplification is often where things break. Anyone deploying AI in public sector environments should read this before writing a single line of code.
002 Graeber · 2015
The Utopia of Rules
David Graeber
Graeber makes a counterintuitive argument: bureaucracy isn't an obstacle to power, it is power. That reframing changed how I approach GovTech. You can't automate your way around institutional logic — you have to understand it first. The most provocative chapter is the one on how bureaucracy creates the illusion of impersonality while concentrating discretion in the hands of those who know the rules best.
003 Christian · 2020
The Alignment Problem
Brian Christian
The best technically grounded book on AI risk that doesn't require a PhD to read. What struck me most was how the alignment problem isn't exotic — it appears in every AI deployment I've worked on, just at smaller scale. When we struggle to specify what we actually want from a system, we've already encountered alignment. Essential reading for anyone responsible for AI governance in regulated environments.
004 Acemoglu & Johnson · 2023
Power and Progress
Daron Acemoglu & Simon Johnson
A serious economic history of technological progress and who actually benefits from it. Acemoglu and Johnson argue that the gains from technology are not distributed automatically — they follow power. The chapter on automation and labor displacement is uncomfortable reading for anyone building AI products. I use this as a counterweight to the default optimism in most AI strategy conversations.
005 Brose · 2020
The Kill Chain
Christian Brose
Brose's central argument — that the West is losing its military-technological edge not because of insufficient R&D but because of how it buys and fields technology — resonates deeply with my experience in the Cyber Innovation Hub. The procurement problem is the problem. The most important defense technology insight in recent years isn't about sensors or algorithms, it's about acquisition cycles. Required reading for anyone working at the defense-technology intersection.
006 Miller · 2022
Chip War
Chris Miller
The best single book for understanding why semiconductor geopolitics is the defining strategic issue of our era. Miller traces the entire history of the chip industry and shows how a technology designed for efficiency became the most contested resource in great power competition. After reading this, every AI strategy conversation I have starts with the supply chain, not the model architecture.
007 Duke · 2018
Thinking in Bets
Annie Duke
Deceptively simple premise: separate the quality of a decision from the quality of its outcome. That distinction is something military planning gets right that corporate environments often don't. Good process can produce bad outcomes; bad process can get lucky. What matters is the decision-making framework, not the result. I recommend this to every product manager and public sector leader I work with.
008 Robson · 2019
The Intelligence Trap
David Robson
Robson documents how high intelligence can become a liability — smarter people are often better at constructing post-hoc justifications for decisions they've already made intuitively. I've seen this pattern in every organisation I've worked in: the most analytically capable people are sometimes the hardest to course-correct. The implications for AI governance, where expert overconfidence is a documented failure mode, are direct and underappreciated.

Updated periodically — last revised 2026

03

Get in Touch

Building AI capability that works in the public sector?

I'm looking for my next role at the intersection of AI and government — in a federal authority, EU institution, or consultancy with a public sector focus. The best way to reach me is via LinkedIn.

Connect on LinkedIn