AI & Automation
AI consultancy for businesses that aren’t sure where to start
Before anyone talks technology, we sit with your team, map how the work actually happens and measure what it costs you. Then we tell you — honestly — what’s worth automating, what isn’t, and what it would take.
- ● Trading since 1995
- ● 4.9-star Google rating
- ● Family-run in Bournemouth
/01 — THE SIGNS A PROCESS IS WORTH A LOOK
The signs a process is worth a look
You don’t need to arrive with a shortlist of technologies, and you certainly don’t need to know what a large language model is. The useful starting point is much simpler: where does your team’s week actually go? In most of the businesses we look after, the candidates announce themselves:
Repeated copying.
Someone re-types the same details from an email into a spreadsheet, then into an invoice, then into a job sheet — several times a day, every day.
Chasing.
Quotes that need a follow-up, invoices that need a nudge, customers who said “call me next week” — all held in one person’s head or on a to-do list that never empties.
Missed enquiries.
Calls and web enquiries that land at 6pm on a Friday and wait until Monday, by which point the customer has phoned someone else.
Spreadsheet glue.
A spreadsheet that quietly holds two systems together, maintained by the one person who understands it — and who occasionally goes on holiday.
The same questions, again.
Opening hours, prices, booking steps, “where are we with my job” — answered by hand, dozens of times a week.
If any of those sound familiar, there’s probably something worth examining. Whether the fix involves AI at all is a separate question — and one we answer with evidence, not enthusiasm.
/02 — HOW DISCOVERY WORKS
How discovery works
Consultancy with us is a structured piece of work, not a sales meeting with slides. It’s the same method we used to automate our own business, and it runs in a fixed order because each step depends on the one before it:
Process interviews.
We sit with the people who actually do the work — not just the owner — and walk through a real week. The exceptions and workarounds they mention in passing are usually where the time goes.
Workflow mapping.
We draw each process as it really runs: every handover, every wait, every “then I copy it into the other system”. You see the map and correct it before we go any further.
Systems and data inventory.
We list the software you use, where your data lives, and which systems have an API — the plumbing that lets one system talk to another. This decides what can be connected cheaply and what can’t.
Baseline measurement.
We count and time the work as it stands today: how many enquiries a week, how long a quote takes, how often invoices get chased. Real numbers, taken from your systems and your people.
Risk and approval mapping.
For each candidate process we agree what a human must check before anything is sent, changed or paid. Consequential actions get an approval step designed in from the start; routine ones don’t need one.
Prioritised opportunity list.
Everything we find, ranked by measured time against build effort — with a plain-English solution outline for the items worth doing first.
/03 — WHY WE MEASURE BEFORE WE PROMISE
Why we measure before we promise
A lot of AI marketing leads with a percentage: hours saved, costs cut, productivity up. We don’t, because without a baseline those numbers are fiction. If nobody has measured how long your invoicing run takes today, any claim about how much faster it could be is a guess dressed up as a promise.
So we measure first. The baseline does three jobs: it tells us whether a process is worth automating at all, it lets us rank opportunities on evidence rather than novelty, and — most importantly — it gives you something to hold us to. If we go on to build a system, you can compare the numbers after go-live with the numbers before, in your own figures, and see whether it earned its keep.
It’s the standard we apply to ourselves. The booking automation, review requests and SMS scheduling that run inside 365 Techies every day were built the same way — look at the numbers first, build second, then check the numbers again. The results are documented on the AI case studies page.
/04 — WHAT YOU GET AT THE END
What you get at the end
Discovery finishes with something you can hold, read and act on — whether or not you act on it with us:
- A map of each process we examined, drawn as it actually runs today.
- Your baseline figures — the counts and timings that describe what each process currently costs you.
- A prioritised opportunity list, with every item honestly labelled: a job for AI, a job for rules-based automation, a process change, or leave it alone.
- A solution outline for the top items — what the system would do, which of your existing tools it connects to, and exactly where a human approves before anything consequential happens.
- A quote — a one-off design and build figure based on complexity, plus the monthly service subscription that keeps the system monitored, maintained and supported once it’s live.
From there the decision is yours: start with the top item, hand us the whole list, or decide the honest answer is “not yet” — the map and the numbers are still yours to keep. And if you already know which process you want examined, you can request an AI quote and go straight to scoping. Prefer to start from your sector? See AI by industry.
/05 — SOMETIMES THE ANSWER ISN'T AI
Sometimes the answer isn’t AI
This is the part of the conversation most AI companies skip. A good share of the problems we’re shown don’t need AI at all — they need deterministic automation: rules-based systems where the same input always produces the same output. If a booking should always create a calendar entry, send a confirmation and add the job to a list, that’s a rule, not a judgement. Rules are cheaper to build, simple to test, and they never have an off day.
AI earns its place where a step needs judgement over messy input — reading an enquiry and drafting a sensible reply, summarising a long email thread, answering a phone call in natural language. In the systems we design, AI does that one step and deterministic automation does everything around it. Most of the working systems on our automation page are exactly this shape, and we’ll always tell you which parts are which.
Sometimes the answer is a process change and no software at all. If discovery shows a form asks for information nobody ever uses, the recommendation is to delete the field — not to build a robot to fill it in. You’ll get that advice just as readily, because advice you can trust is the entire point of asking for it.
/06 — ADVICE FROM PEOPLE WHO RUN THESE SYSTEMS EVERY DAY
Advice from people who run these systems every day
There’s no shortage of AI consultants at the moment. What’s rarer is one who can show you the things they’re describing actually running in their own business. We can. The booking automation, review-request emails, SMS scheduling and Slack-connected workflows we recommend are the same ones running inside 365 Techies today — built in-house, in production, doing real work. Our 365 AI OS demo shows a Claude-powered AI desktop we built ourselves (a scripted demonstration, and clearly labelled as one), and our monitoring dashboards run live on this site.
We’re not new to unglamorous, reliable systems either. We’ve been a family-run IT support business in Bournemouth since 1995, we hold a 4.9-star Google rating, we’re a Microsoft Partner, and our telemetry work earned Victron Energy’s Recommended Software Integrator status. Whatever comes out of your discovery is supported by the same local team that looks after your IT — across Bournemouth, Poole, Christchurch and the wider Dorset area, and remotely across the UK where that suits. If you’d like to see what the builds themselves look like, start with our AI services overview.
Open the 365 AI OS demo// SERVICE & BUILD
Monthly service — confirmed with your quote
One-off design & build — quoted by complexity Consultancy and discovery work is scoped and quoted before it starts. A monthly service subscription applies where we go on to implement a system for you — it covers monitoring, maintenance and support from the same local team that looks after your IT.
// GOOD QUESTIONS
Frequently asked
We don’t know anything about AI. Is that a problem?
No — it’s the normal starting point, and it’s why this service exists. You never need to pick a technology: you describe the work, we map it, and we come back with recommendations in plain English. Where a term matters — an API, a large language model, deterministic automation — we explain it the first time we use it. The decisions you’ll actually be asked to make are business decisions: which process matters most, what a human must approve, and whether the quote stacks up against the measured time that process costs you.
How long does discovery take?
It depends on scope, which is why it’s quoted rather than sold off a rate card. Looking at one process — say, how enquiries are handled from first contact to booked job — is a much smaller piece of work than mapping a whole business. When we scope the work we’ll tell you how many interviews we need, whose time we’ll take and for how long, and when you’ll have the written findings. Most of it fits around your team’s normal day; nobody has to down tools for a week.
What does AI consultancy cost?
Consultancy is scoped and quoted before any work starts, based on how many processes and systems we’re examining — so you’ll know the figure up front. If the findings lead to a build, that’s quoted separately by complexity, and the finished system then runs as a monthly service subscription covering monitoring, maintenance and support. There are no tiers to decode and no obligation to build anything: the findings and the baseline figures are yours either way.
Won’t you just recommend AI because AI is what you sell?
No — and the written findings make that hard to fudge, because every opportunity is labelled as AI, rules-based automation, a process change, or not worth doing. In our own business a large share of the wins turned out to be plain rules-based automation: cheaper to build, simpler to test, with AI reserved for the steps that need judgement over messy input. We make our living from ongoing service on systems that keep working, so recommending the wrong tool would cost us more than it earned.
What happens to our data during discovery?
Discovery is mostly conversation and observation — we map how work flows and where data lives, and we don’t need copies of your customer records to do that. We never ask for your passwords. Where a recommended system would handle personal data, the solution outline states exactly what it touches, which systems it connects to, and where a human approves before anything is sent or changed. Data protection is treated as a design constraint from the first drawing, not a box ticked at the end.
Do you only work with businesses near Bournemouth?
We’re based in Bournemouth, and most of our clients are across Bournemouth, Poole, Christchurch and the wider Dorset area — which is where on-site interviews are easiest to arrange. But discovery works well remotely too, with interviews over video call and the workflow maps shared on screen, so we take on work across the UK where that suits. What doesn’t change is the accountability: a named, family-run team that’s been trading since 1995, and that you can phone.
Not sure where to start? That’s the point of this.
We’ll map the process, measure the baseline and give you a prioritised, honest plan — including when the answer isn’t AI at all.
01202 775566 · help@365techies.co.uk · MON–FRI 9AM–5PM