AI Workflow Design · London & UK

Your team knows the tools.
The work hasn't changed.

AI workflow design and automation training for teams who have already had the training.

Most organisations have now bought the licences and run the training. The tools changed. The work didn't. AI workflow design is the layer that closes that gap: we take the jobs your team actually repeats every week and rebuild them as AI workflows that run reliably, in the stack you already pay for.

Beatrix Meszaros guiding a participant through an AI workflow on his laptop during a hands-on workshop

AI workflow design sits one level above tool training. Instead of teaching people what a tool can do, these sessions redesign a specific piece of recurring work and hand it back automated.

Whether you call it AI workflow automation, AI process automation or building agents, the outcome is the same: fewer repeated requests typed by hand, more work that runs on its own. Delivered in person in London and across the UK, or online for distributed teams, in English or Hungarian.

The gap

Trained on the tools,
still doing the work by hand.

01

Everything is still a one-off

People open a chat window and retype a version of the same request they made last week. It works, but nothing accumulates. Every good result is thrown away the moment the tab closes.

02

The good stuff stays with one person

One person on the team gets very good at it. Nobody else can reuse what they built, because it lives in their head and their chat history rather than in something the team can run.

03

Nothing runs unless somebody starts it

The weekly report still waits for someone to remember to make it. Work that could trigger itself is still queued behind a human deciding to begin, which is where most of the promised time saving quietly disappears.

How it works

Three steps, and the first one
happens before I arrive.

Every engagement starts with a short discovery call to establish which AI tools you actually have access to. Sessions are then built inside that stack, within whatever your IT and governance rules allow.

01

Audit the stack you already have

Which AI tools are licensed, who can use them, what the governance rules say, and what is already quietly in use that IT does not know about. This determines everything that follows.

02

Map the work worth changing

The team lists what they actually repeat every week. We sort it by how often it happens and how mechanical it is, and pick the handful where automation returns real hours rather than novelty.

03

Build the workflows, in the room

Reusable instructions the whole team can run, connected to the systems where the work lives, with the repeating ones set to run on a schedule or a trigger. Built by the team, not demonstrated at them.

The formats

Three ways to run it,
depending on where you're starting.

Pricing is built around your team size, the format and whether delivery is in person or online. Tell me where your team is and I will tell you which of these actually fits.

Abstract illustration of scattered shapes being sorted into ordered clusters, representing AI workflow discovery
Half day · 8–20 people

Workflow Discovery Workshop

Best if your team is convinced AI should be saving time somewhere, but nobody agrees on where to start.

  • A full audit of the AI stack you are already paying for
  • The team maps the work they genuinely repeat every week
  • Everything is scored and prioritised by hours returned
  • Live demonstrations of what is realistically possible in your stack
You leave withA workflow map of your team's real work and a prioritised, costed plan of what to build first.
Abstract illustration of modular blocks assembling into a stable structure, representing an AI workflow build day
Full day · plus follow-up

Workflow Build Day

Best if your team already knows what it wants automated and needs a day with someone who can actually build it with them.

  • Reusable skills the whole team can run, not one person's prompts
  • Scheduled and triggered runs, so work happens without being started
  • Connected to the systems your team already works in
  • A follow-up session 30 days later to fix what drifted
You leave withA working set of team-wide AI workflows, documentation for each, and a review booked to make sure they survive.
Beatrix Meszaros leading a hands-on AI workshop with a team working on laptops
The deliverables

Built in the room,
yours when you leave.

Nothing here is a slide deck you never open again. Everything is built live, on your work, in your tools.

Working AI workflows, running in your own stack
Reusable skills the whole team can run
Scheduled runs for the work that repeats
A one-page map of what was automated and why
Written documentation for each workflow built
A named owner for every workflow, so it survives
For leadership

The question isn't whether
people are using AI.

By now, most of them are. The harder question is whether any of it has changed how the organisation operates, or whether you are funding a lot of individually impressive activity that never reaches the P&L.

  • Time returned is measured against named workflows, not self-reported confidence
  • Every workflow has an owner, so adoption outlives the enthusiasm of one champion
  • Built inside your existing licences, governance and data rules, with nothing new to procure
  • A clear view of which teams are ready to go further and which need the groundwork first
Beatrix Meszaros presenting the four stages of AI maturity to a corporate audience
For small businesses and solo operators

No team to delegate to?
That's exactly the point.

When you run a small business, the repetitive work does not get delegated, it gets done by you at the end of the day. The same method applies, scaled down: we look at where your week actually goes, then build the handful of AI workflows that take the most tedious parts off your desk permanently.

Delivered remotely as a focused one-to-one session, or as a small cohort if you would rather build alongside a few other owners. No enterprise stack required, and no assumption that you have an IT department.

Enquire about a one-to-one
Not sure which format fits

Find out where your team actually is first.

The AI Maturity Assessment shows which dimensions are holding your team back — awareness, tool adoption, workflow integration, measurement, culture. It is the fastest way to know whether you need to map the work or start building it.

Take the assessment → 5 minutes · No sign-up required
AIREADY
Awareness
Tool adoption
Workflow
Measurement
Culture
FAQ

Questions about
AI workflow design.

It means redesigning a piece of recurring work so AI handles the repeatable parts reliably every time, rather than someone opening a chatbot and retyping the same request. In practice that means building reusable instructions the whole team can run, connecting them to the systems where the work already lives, and setting the ones that repeat to run on a schedule or a trigger.

The training courses teach people how to use the tools. Workflow design starts after that and changes the work itself. If your team has already had ChatGPT training or Microsoft Copilot training and you are still not seeing the time come back, this is the layer that is missing.

Whichever ones you already pay for. Every engagement starts with a short audit of the AI stack you have licensed and what your IT and governance rules allow. Sessions are then built in that stack, whether that is Microsoft Copilot, ChatGPT, Claude, Gemini or a combination of them. Nobody leaves the room having been sold six new subscriptions.

No. Everything is built without code. The people who should be in the room are the ones who actually do the work being redesigned, because they are the only ones who know where the time really goes.

Workflow sessions work best with 8 to 20 people, because everyone builds rather than watches. Larger groups can be split across repeated sessions on the same day, and I will tell you honestly if a group size will undermine the hands-on part rather than take the booking regardless.

The workflows built in the room work in the room, so the first savings land the same day. The more useful question is whether they are still running a month later, which is why the full day includes a follow-up session 30 days on to fix what drifted and rebuild what nobody used.

Yes. The same method works for one person or a team of three, delivered remotely as a one-to-one session or a small cohort rather than an on-site workshop day. Small businesses often see the biggest change, because there is nobody to delegate the repetitive work to.

Yes. Sessions run in person across London and the UK, on site at your office, and online for distributed teams. In-person delivery works best for the hands-on building, and sessions are available in English or Hungarian.

Ready to start?

Let's find out what your team
should stop doing by hand.

Tell me your team size, which AI tools you already have and what keeps eating the week. I'll come back with the format that fits and what it would cost.

What participants say

Hands-on sessions
that people come back to.

The practical approach with live examples. Learning the different prompt types — nothing I would change.

AI Training participant · Asset Management Firm, London

I am just not missing her sessions. Practical as always.

AI Training participant · Brainster

Every session with Beatrix is amazing, content relevant and easy to follow. I simply soak up knowledge!

AI Training participant · Brainster