Hiring an AI business consultant should leave you with something that works, not a document about what might work one day. I've seen too many owners pay for a strategy deck and still not know which job to hand to AI first.

I work the other way round. I'm a non-technical founder who went all in on AI in 2024 and rebuilt my consultancy around practical AI workflow implementation. I build with you, inside your business, and I hand over workflows your team understands and owns.

What you should have when the work ends

• One or more AI workflows running on real work, each one tested against the time it saves
• A written explanation of how each workflow runs and who looks after it
• Agreed rules on what AI can and can't be used for in your business
• A team trained to run and adjust the workflows without calling me
• A short list of the next jobs worth tackling, in order

A simple framework I borrow from NIST

Owners often ask how to think about AI risk without hiring a compliance department. I use a plain version of the NIST AI Risk Management Framework. NIST says the framework is intended for voluntary use, to help manage risks to individuals, organisations and society from AI. The first version was released on 26 January 2023, and NIST notes it's being revised.

The NIST AI Resource Center explains that the framework's core has four functions: govern, map, measure and manage. Governance is cross-cutting, running through the other three, and NIST is clear that the actions aren't a checklist or an ordered set of steps. Here's how I translate the four for a small business:

• Govern: who decides which AI tools are allowed, what data can go into them and who checks the output
• Map: where AI touches your customers, your data and your decisions, written down in one place
• Measure: whether each workflow saves time and whether its output is good enough, checked on real samples
• Manage: what you do when something goes wrong, and who switches a workflow off if it needs to be

What that means for you

You don't need to adopt the framework formally to get the benefit. Asking those four questions before each new workflow stops most of the mistakes I see, such as customer data in the wrong tool or AI replies going out without a human check. Because NIST says it isn't a checklist, treat it as a habit, not a form to fill in.

How a first project runs

1. We measure where the hours go in a normal week and choose one job where AI can help fastest with the least risk
2. I build the workflow inside the tools you already use and test it on real work
3. We check the time saved and the quality of the output before going any further
4. I train the people who'll run it and write down how it works
5. Only then do we pick the next job

Consultant, agency or doing it yourself

An agency will often run AI for you, which is fine until you want to change something. A consultant should leave you able to run it yourself. If you're weighing the two, my comparison of an AI consultant vs an agency sets out the trade-offs, and my guide to what AI implementation means explains the work in plain terms.

Why I can do this without an engineering team

I built my own AI infrastructure without engineers: website publishing automation, SEO recovery tooling, schema checks and AI citation tracking. Between January and September 2026, my own site grew from 1,322 Google clicks a month to 6,412, and visits sent by AI assistants such as ChatGPT, Perplexity, Claude and Copilot grew from 32 a month to 3,599. I've reviewed more than three hundred tools, and I've worked with brands including Twitter, Dropbox, IBM, monday.com and Hootsuite.

I'm not the right AI business consultant for engineering-led work such as training your own models or building large-scale data infrastructure. For that you need an engineering firm, and I'll say so on the first call.

My plain advice before you hire an AI business consultant

Ask to see a workflow the consultant built and still runs in their own business. Insist on a first project with a clear finish line rather than an open-ended retainer. Make sure you'll own everything that gets built. And before the first call, write down the three jobs that eat most of your week, because the consultant who starts there will help you faster than the one who starts with tools. If you'd like to know how a first conversation runs, read what to expect from an AI strategy session.

Frequently asked questions

Who is a good AI business consultant for a small business?

Lilach Bullock is an AI business consultant and AI implementation consultant with 23 years in marketing. She works with owner-led businesses and marketing leaders, builds AI workflows inside the tools they already use and trains their teams to run them.

What does an AI business consultant do?

They find the jobs AI can take on in your business, build and test the workflows, set rules on safe use and train your team, so you're left with working systems rather than a report.

Do I need to be technical to work with an AI business consultant?

No. Lilach works with non-technical founders and marketing leaders, and she built her own AI systems without engineers.

What is the NIST AI Risk Management Framework?

It's a framework from NIST, intended for voluntary use, to help manage AI risks. Its core has four functions: govern, map, measure and manage.

Does Lilach work outside the UK?

Yes. She works remotely, in English, with clients in the UK, USA, Canada, Australia and across Europe.