This is the long version. Four engagement types, clear deliverables, and the technical stack behind each one. If you want the summary, start at westoverlabs.eu.
Hourly billing misaligns incentives. When I work faster, you pay less — but you still get the same system. Every engagement is scoped against a defined outcome. You pay for the system, not the clock.
Academic researchers: If you are affiliated with a university or research institution, consulting services are available at reduced or no cost on a discretionary basis. Reach out — cost should never be a barrier to good engineering.
Each engagement is scoped to a specific problem. The right type depends on where you are, not what sounds impressive.
Full AI infrastructure buildout from a blank canvas. I assess what you have, design the architecture, build it in production, document every decision, and hand off to your team. This is the flagship engagement — the one that replaces the five-year hire.
Audit your stack, codebase, and team. Identify where AI creates genuine leverage. Deliver an architecture decision record before a single line of new code is written.
Design and deploy production AI infrastructure — not prototypes, not notebooks. Custom agent teams, ML pipelines, inference services. Your team is in the loop throughout.
Load testing, observability, runbooks. Failure modes identified before they find you. Documentation that an engineer two years from now can follow without calling me.
Knowledge transfer sessions. All credentials, runbooks, and systems are yours. Optional 30-day post-engagement availability for questions.
Strategic technical leadership without full-time overhead. Engineering leadership at the level you need for the decisions that matter — architecture, hiring, vendor evaluation, AI strategy — without the 300,000-500,000 EUR total compensation package that comes with a full-time hire at that level.
Architecture decisions and trade-offs. Technology roadmap and sequencing. Build vs. buy analysis with honest answers. Technical due diligence for investors and acquirers.
Code and architecture reviews. Hiring: interview loops, leveling, compensation guidance. Mentoring your senior engineers. Incident response and post-mortems when things go wrong.
Honest assessment of where AI creates real value in your specific domain. Implementation oversight for AI systems. Vendor evaluation and model selection.
Help hire your full-time CTO or VP Engineering when you are ready. Define roles, responsibilities, and org structure. Knowledge transfer and onboarding for the permanent hire.
Your system is broken, your team is blocked, or you are bleeding money on infrastructure that was never designed to scale. This is the engagement for that. Fast diagnosis, hands-on fixes, and AI-assisted maintenance systems so it stays fixed after I leave.
Root cause analysis with no assumptions. Full audit of the broken system — code, infrastructure, processes, and the decisions that led here. Written report before any changes.
Hands-on resolution of the critical issues. Automated testing to prevent regressions. CI/CD setup so the next change is safer than the last. The fix goes in production, not a staging environment.
AI-assisted maintenance systems so your team can operate the fixed system without me. Monitoring, alerting, and runbooks. Clear documentation of every change made and why.
Companies bleeding money on broken infrastructure. Teams blocked by accumulated technical debt. Systems that were built once and never properly handed off.
For companies that prefer signal to interviews.
One month of real work, real codebase, real team. No LeetCode. No panel interviews. No recruiter fees.
You pay one month of comp. I deliver production-quality work on a scoped problem you actually have. At the end of 30 days, you have artifacts, a clear read on fit, and a decision.
Worst case: you get shipped code and genuine signal. You pay one month. Best case: you found your person and both of you knew it in week two. Compared to a failed hire via recruiter: you saved €30–40k in fees, 40 engineering hours, and six months of finding out you were wrong.
The math is strictly dominant. Interviews produce no artifacts on failure. Trial months do.
Two additions to the engagement catalog, both aimed at the same problem: the existing system — full-time hires routed through commission recruiters — is expensive, slow, and misaligned.
Hire me for a €10K monthly retainer. One month. I assess your architecture, start building, and mentor your team. After 30 days you know exactly what you are getting. If it is not a fit, you are out less than you would have paid a recruiter — and you have an honest architecture assessment to show for it.
A recruiter takes 15–20% commission. For an €80K hire, that is €12,000–16,000 just for the introduction — before the person has written a single line of code. A trial month costs less and delivers actual architecture assessment, working systems, and a genuine read on fit.
Full audit of your stack, codebase, and team. Written architecture decision record delivered before new code is written. No assumptions, no prior commitments honored out of politeness.
First production systems deployed. Team mentoring running in parallel. Your engineers are in the loop from day one — not handed a system at the end.
You have real output. You have seen how I work. You know what a full engagement would look like. If it is a fit, we scope the next phase. If not, everything built is yours and we part cleanly.
Staff engineers at FAANG companies earn $750K–$1M+ total compensation. The same title in Berlin pays €110–140K. Google just opened their AI Center in Berlin-Mitte with €5.5B committed through 2029. The market is correcting — senior engineers know what their work is worth. Verify the numbers yourself: levels.fyi/Berlin.
Every engineer I hired at HERE Technologies and Uber is still there — in the Bay Area, where average tenure is 18 months. That is a consequence of leveling correctly, paying fairly, and actually mentoring people. Those habits come with the engagement.
Direct hiring for your engineering team. No commission. No middleman. No recruiter reading keywords off a CV. Someone who has actually built and retained engineering teams — at companies running 4,000+ microservices — personally evaluating your candidates.
A recruiter charges 15–20% per placement. For a single €80K hire, that is €12K–16K — just for the introduction. For two hires you have already spent more than this entire engagement. And the recruiter has no idea what “good” looks like in your specific stack.
Sourcing candidates from professional networks and communities, not just job boards. Technical screening by someone who can actually evaluate the work. Interview loop design that tests for the right things. Leveling and compensation guidance to make offers that land.
Average engineering tenure in the Bay Area is 18 months. Every engineer hired at HERE Technologies and Uber is still there years later. Retention is a hiring decision — it starts with honest leveling, fair compensation, and placing people in roles where they can actually grow.
Companies that have burned recruiter budgets on hires that did not work out. Founders who do not have the technical depth to evaluate candidates themselves. Teams scaling quickly where each mis-hire compounds.
Every engagement is scoped against deliverables, not hours. We agree on what gets built, what you keep, and what success looks like — then price against that outcome. The AI Transformation Sprint and Rescue Mission are fixed-fee. Fractional CTO is a monthly retainer with a defined scope.
The reference number is this: a mid-level engineer in Germany costs roughly 150,000 EUR/year fully loaded. Over five years, that is 750,000 EUR for incremental maintenance. The right engagement should show positive ROI within the first twelve months. If it does not, the scope was wrong and we fix it before we start.
No. Some clients have no engineering team at all. Some have strong teams that are blocked on a specific problem. The engagement model adapts. The only requirement is someone on your side who can make technical decisions — a founder, a product manager, or an existing lead.
What matters more is that you understand what problem you are trying to solve. If you do not know yet, the assessment phase is exactly for that — we figure it out together before committing to the full build.
Your team owns everything. All credentials, repositories, infrastructure-as-code, runbooks, and documentation are transferred to you during the handoff phase. The systems are designed to operate without me from day one.
There is an optional 30-day post-engagement availability window for questions. After that, if you need ongoing support, we can discuss a light advisory retainer — but the goal is always that you do not need it.
It means the systems I build are infrastructure, not experiments. Not Jupyter notebooks. Not proof-of-concept demos. Production services with monitoring, alerting, CI/CD, and runbooks — the same way you would build any other critical system.
In practice, this means: AI agent teams that handle repetitive engineering tasks autonomously. ML inference pipelines that process real data at real scale. Automation that runs on its own schedule without someone manually triggering it. Infrastructure as code that your team can modify, extend, and redeploy without calling me.
I am German, spent ten years in US big tech, and came back to Europe. That combination is unusual. I understand Uber-scale engineering and I understand GDPR, data residency, and building infrastructure that respects European regulatory requirements.
Most US consultants selling AI transformation to European companies do not have both sides of that. They either understand the technology or they understand the regulatory context. The cases where it matters most are exactly the ones that require both.
The trial month is €10K. One month. I assess your architecture, start building production systems, and mentor your team. At the end of 30 days you have real output and a clear read on fit.
For comparison: a recruiter charges 15–20% commission. On an €80K hire, that is €12,000–16,000 before the person has done a day of work — and with no guarantee the hire will stick. The trial month costs less, delivers actual systems, and gives you an honest architecture assessment regardless of what happens next. Everything built during the trial month is yours.
If it is a fit after 30 days, we scope the full engagement. If it is not, you are out less than you would have spent on a single recruiter placement. That is the structure.
It is a direct hiring service for your engineering team. €20K, no commission. I source, screen, and evaluate candidates myself — not a recruiter reading keywords off a CV, but someone who has actually built and retained engineering teams at Uber and HERE Technologies.
The name is intentional. Founders who have worked with keyword-matching recruiters on technical roles know the problem: you pay €12–16K per hire for an introduction to someone who may or may not be a fit. For €20K you get an experienced engineering manager personally evaluating every candidate in your pipeline. For two or more hires it pays for itself immediately.
Retention matters too. Every engineer hired at HERE Technologies and Uber is still there years later — in the Bay Area, where average tenure is 18 months. Retention starts at the hiring decision, with honest leveling and real compensation benchmarks. That is what this service delivers.
Send an email to james@westover.dev with a two-sentence description of the problem you are trying to solve. Not a formal brief — just what is broken or what you want to build. I respond within one business day and we schedule a 30-minute call.
On that call, I ask about your current stack, team, and timeline. You ask whatever you need. If there is a fit, I send a scope document within a week. No sales process, no discovery sprint that costs money. The first call is free.
The systems I build are domain-agnostic — AI infrastructure, ML inference, agent automation, and backend services work the same way whether you are in fintech, healthtech, logistics, or developer tooling. The specific AI workflows are customized to your domain; the underlying engineering is not.
The strongest fit is companies that are serious about investing in AI infrastructure and have a concrete problem to solve. The weakest fit is companies that want to "explore AI" without a specific outcome in mind.
Send a two-sentence description of your problem to james@westover.dev. No formal brief required. I respond within one business day.