Services
Generative AI consulting
Generative AI consulting usually sells you on what a language model can do. Half of the job is telling you where it can’t, and where a plainer tool does the same work for less money and less risk.
When a language model is the wrong tool
Some problems don’t need a language model. Part of the job is telling you which ones. A board that has asked you to “do genAI” wants a number and a plan, not a demo. So the first thing we do is separate the use cases where a model pays off, usually messy language work like drafting, summarizing, triage, and search over your own documents, from the ones where a rule or a small classic model is cheaper and more predictable. You leave the assessment knowing which is which, with costs attached.
How an engagement works
Everything starts with the assessment: fixed scope, 3–6 weeks, priced at $20,000 to $80,000 (where in that range depends on scope: how many systems and teams we assess, company size, and regulatory exposure). At the end you get a concrete recommendation: what to build, what to buy, what to leave alone, and what it costs to run safely. Larger work only follows when the assessment has earned it, and it’s fixed scope too. For context, published 2026 ranges put a typical scoped production build at $50,000–$250,000. And that band is a single scoped build; multi-system, multi-year programs run seven figures and beyond, and scale like that is work we take.
What you get out of it
- A short list of generative AI use cases ranked by payoff, and an honest note on the ones that aren’t worth it.
- A working system in production on your own infrastructure, with the retrieval, prompts, and evaluations documented so your team can run it.
- Guardrails your auditors and your lawyers can live with: access limits, logging, and checks on what the model is allowed to say and do.
- A cost you knew before the work began.
Where this connects
Once the assessment picks the work, we build it into production. If the valuable use case turns out to be an autonomous one, that’s AI agent consulting. Running any of it safely is AI governance consulting, and the full menu is on the services page.
Who you’d be working with
Nick Major, an engineer of fifteen-plus years who builds AI systems in production, and Isaac Major, an operator who has run growth and revenue teams inside global enterprises. They work from Lisbon and Salt Lake City, which between them covers most of your working day. Full backgrounds are on the about page. We’re a young firm, so there are no case studies or client logos here yet; what we show instead is published pricing, a fixed-scope process, and the senior people who do the work.
Questions people ask
- What does a generative AI consultant actually do?
- Four things, in order: find the handful of places where a language model earns its cost, evaluate whether an off-the-shelf model plus retrieval beats a custom build, ship it into production on your infrastructure, and put guardrails around it so it can run without babysitting. A good one also names the places where generative AI is the wrong answer.
- When is generative AI the wrong tool?
- Often. If the task has a correct answer that a rule or a small classic model can produce, a language model is a slower, pricier, less predictable way to get it, and it can be wrong with confidence. We flag those cases in the assessment instead of billing you to build them.
- How much does generative AI consulting cost?
- Our engagements start with a fixed-scope assessment at $20,000 to $80,000, delivered in 3–6 weeks. Published 2026 ranges put a scoped production build at $50,000–$250,000. The full market picture, including hourly and retainer ranges, is in our AI consulting rates guide.
- Who owns the models and prompts you build?
- You do. We build on your infrastructure and hand over the code, the prompts, the evaluations, and the documentation. We take no resale margin and no vendor commission, so the model recommendation is the one that fits your problem, not our margin.
- Why are there no case studies here yet?
- Because we are a young firm and we will not invent them. What we can show today is real: published pricing, a fixed-scope process, and the two named founders who do the work. Case studies appear when clients agree to be named, and not before.
Tell us about the work.
A few lines is enough. We read every enquiry ourselves and reply within one business day.