AI consulting
that ships code.
Independent AI consulting for businesses building production AI in regulated environments. Six engagement shapes ordered by commitment — start with a €750 Technical Assessment or a €8–10K Scoping Engagement, scale into a longer build if it's the right fit. No fixed packages: shape gets decided once we've named the actual problem.
Six ways
to work together.
Ordered from lightest entry to deepest commitment. Most teams enter at the Technical Assessment or Scoping Engagement and decide from there.
1 · Technical Assessment — €750 · 90 min · paid working session
Lightest entry. Bring an architecture diagram, a vendor proposal, an eval question, or a specific stuck point. I review it live, ask the questions a senior consultant would ask, and send a one-page written take within 48 hours.
Useful for validating fit before a longer engagement, or just getting senior eyes on a decision your team can't agree on.
2 · Scoping Engagement — €8–10K · 2 weeks · fixed price
Two weeks of front-loaded diagnostic work producing concrete artifacts your team, leadership, and Legal can act on: stakeholder interviews, architecture document, sovereignty / vendor decision record, realistic build plan with phasing, cost estimate. End of week 2: a written readout and a meeting.
Best for teams facing a build/buy decision on a meaningful AI investment, where the cost of getting the architecture wrong is high. Deliverables stand on their own — even if you don't continue working with me, you walk away with a usable plan.
3 · Advisor Retainer — €7–11K/month · 3-month minimum
Weekly architecture call (90 min) plus async availability between calls. Your team owns delivery; this engagement keeps them aimed at the right target. Architecture decision support, vendor management input, eval discipline, code-review feedback as needed. No implementation — your team writes the code.
Hours scale with price: €7K/mo covers 5–6 hrs/wk (call + async); €9K/mo adds light hands-on review; €11K/mo includes active design work and code review.
4 · Architecture & Audit — 1–3 months · part time · quoted on call
I review your AI stack, surface the risks (vendor lock-in, scaling cliffs, missing evals, observability gaps), and deliver a written migration or improvement plan with options and trade-offs. I can lead the implementation or hand it off to your team.
This was the engagement at Roche — what started as a LangChain → LangGraph migration ended with a vendor lock-in fix in the observability layer that turned out to be the bigger risk.
5 · AI Product Development (embedded build) — 3–6 months · €15–22K/mo
I embed with your product team and own AI features end-to-end: architecture, implementation, evaluation, deployment. RAG copilots, AI agents, meeting bots, structured-output features, LLM automation. I've shipped this work inside SaaS platforms (Teamwork.com), enterprise tools (Roche), and automation systems (Universal Music Group).
Monthly retainer scoped on the discovery call. Engagements start at three months — less than that and there isn't enough time to do useful work. Part- or full-time depending on the scope.
6 · AI Team Setup — 3–4 months · 10–15 h/week · quoted on call
I hire your first AI engineers, define the tech stack, plan the architecture, and mentor the team until they can ship without me. Then I step back. The goal is your independence on a schedule.
This is what the EuroPharma engagement looked like — three months from "no AI team" to a team that ran their own roadmap.
How I price.
Technical Assessment and Scoping Engagement are fixed-price — €750 and €8–10K respectively. Advisor Retainer is published above (€7–11K/mo, scales with hours). Embedded Build runs €15–22K/mo for a typical full-time engagement; audit and team-setup engagements are quoted per scope on the discovery call.
I don't quote hourly rates on a public page because the answer depends on what we're solving. For reference, senior AI consulting in the regulated / production-AI space typically lands between €120–180/hr — happy to walk through specifics on the call.
Two-week notice either side. If the engagement isn't earning its keep, we end it cleanly. The work I've done is yours — in your repo, with docs.
What the first
month looks like.
No two-week pitch deck. We start with a 30-minute call, agree on the shape of the work, and I'm pushing code by week two.
Discovery call
30 minutes. You describe the problem; I ask the questions a senior consultant would ask. No pitch.
Read the code
I review your codebase, talk to your engineers, and write a short plan with options and trade-offs.
First PR
Embed in your repo and standups. Ship something small but real — a fix, a benchmark, a working spike.
Own the outcome
Take the AI feature end-to-end. Deploy, monitor, iterate. Hand over with docs and a team that can run it.
Who this is
actually for.
Companies adding production AI to an existing product, or building one from scratch in a regulated environment (pharma, finance, healthcare-adjacent SaaS, regulatory intelligence). Teams that have raised seed or Series A and need senior AI engineering judgment without the time-to-hire of a full-time lead.
I work well with technical founders, engineering directors, and CTOs who already have a strong eng team and need an AI specialist who can write the code, not just review it. I've shipped features through Roche's IT review and built systems where Legal sign-off was the hard requirement, not a nice-to-have.
I'm probably not a fit if you need an offshore contractor at €50/hr, a generic "AI strategy" consultant who doesn't write code, or a 12-month full-time hire. Those are real needs — just not what I do.
Ready to talk?
30-minute discovery call. I learn your problem, you learn if I'm the right fit. No pitch, no commitment.
Book a free call → or email maciek@studiochrome.ai