Hire meAI data readiness, private AI implementation, training.
Private AI consultant: I help teams get their data ready for AI and run it on their own machines.
I am Frank Vitetta, an AI implementation consultant based in London. I work as a private AI consultant for teams that want to use AI without sending their code or data to someone else's cloud. We get the data ready first. Then we run AI on your own machines where that matters.
Services
Three services, from an AI data readiness audit to training
-
N° 01
AI data readiness audit and strategy
I look at what data your team has and how good it is. I measure it against the standard data quality dimensions. Those are completeness, uniqueness, timeliness, validity, accuracy and consistency. Then I set out what an AI or retrieval project would need from that data. I also say which parts of the work could run on local models.
What you getA written assessment and a prioritised plan. It is a data strategy for AI that says what to fix first and why.
-
N° 02
Private AI implementation
I set up AI that runs on your team's own machines or under your own keys. I use existing tools. That means local model runners such as Ollama, LM Studio and llama.cpp, plus Code Scout where it fits. I work alongside your developers and analysts, so the setup fits how the team already works. This covers local LLM implementation and AI coding agents for teams.
What you getA setup on your own hardware or accounts. Written notes on how it is put together, so your team can run it.
-
N° 03
Training and workshops
Hands-on sessions for developers, analysts and their managers. We cover working within the limits of local models. We also cover reading benchmark claims with care and keeping data where it belongs.
What you getA session planned around your team's own tools and questions. You keep the materials.
Disclosure. Code Scout is my own project. Tool choice follows the audit and is not tied to my products.
Process
How the work runs
The order is the same each time. Every step after the first is covered by a written, fixed-price proposal before it starts.
- Working sessionOne paid session on your actual problem, at a fixed price quoted in writing before you book.
- AuditWhat data you have, how good it is and what could run on local models.
- PlanA prioritised plan and a written fixed-price proposal before any further work starts.
- ImplementationPrivate AI set up with existing tools, alongside your developers and analysts.
- TrainingHands-on sessions, so the team can carry on without me.
Fit
Who it suits
- Teams that want to use AI without sending their code or data to someone else's cloud. Some cannot. Others would rather not.
- Development teams weighing AI coding agents and local models against hosted subscriptions.
- Marketing and analytics leaders preparing their data for an AI or retrieval project.
- Managers who have to judge AI and benchmark claims without an engineering background.
Limits
What I do not do
I do not build data pipelines or offer data engineering. I do not train or build custom machine learning models. I set up existing models and tools. I come from the marketing data side. What I bring is explanation, the right questions and judgement, with sources.
I do not give legal advice either. If what you need is a data engineer, a machine learning engineer or a lawyer, I will say so.
Background
Why me
I am based in London and have worked in search marketing since 2008. I founded Code Scout. It is a free desktop AI coding agent that can run local models, so code stays on the user's machine.
My specialisms are technical SEO, structured data, site architecture and AI readiness. I help businesses adapt as people move from search engines to AI assistants to find things.
The public sources for all of this are listed on the about page. The field notes show how I explain things.
Questions
Questions about hiring a private AI consultant
How much does it cost?
I do not publish prices. The first step is one paid working session at a fixed price. I quote it in writing before you book. Any work after that is covered by a written fixed-price proposal. You receive it before the work starts.
How long does it take?
That depends on the scope. The timing is set out in the written proposal, so you see it before you agree to anything.
Which tools do you use?
I choose tools after the audit. For local models I work with existing runners such as Ollama, LM Studio and llama.cpp. I may suggest Code Scout if an AI coding agent fits. Code Scout is my own project. I say so whenever a tool of that kind comes up. Tool choice is not tied to my products.
What happens to our data?
The aim of private AI is that your code and data stay on your own machines or under your own keys. We agree in the written proposal how any data is handled during the work. That happens before the work starts. The privacy page covers what you send through the form on this page.
Do you guarantee results?
No. I do not guarantee outcomes. Local models have limits. Some work is a poor fit for them. The audit sets out which work could run on local models. I will say clearly when they are not the right choice.
How do we start?
Use the form on this page to ask about a working session. Tell me what you are trying to do and what you have in place. I reply by email with a written fixed price for the session. Then you decide whether to book.
First step
Ask about a working session
The first step is a paid working session. It is one fixed-price session on your team's actual problem. I quote for it in writing before you book. It covers:
- The problem you want to solve and who else is involved in it.
- What data and tools you already have, described in your own words.
- Which of the three services, if any, is the right next step.
- Whether I am the right person. If I am not, I will tell you.
Tell me a little about the problem and what you have in place.
Disclosure. Enquiries come to me personally and I answer them myself.