The Meridian Company

Know where AI pays off in your business. Build it before your competitors do.

Meridian is the AI transformation partner for companies whose growth leaves things better: cleaner power, smarter factories, healthier food.

Apply for a blueprint sessionSee how it works

Clients

EdenLabPepsiCoJewish Climate Trust

The team has taught AI inside

Octopus Energy, Google, BlackRock and NHS Wales, and to 10,000+ energy, industrial and natural capital professionals.

The blueprint session

Two weeks inside your operations. Every place AI pays back, scored and ranked, with the top pick specced ready to build.

Time to your map
2 weeks
Time from your leaders
3 hours, total
If you don’t judge it worth 10× the fee
You don’t pay

AI is redrawing the map on competitive advantage.

Every market is about to be re-ranked. The companies that come out ahead will be the ones that picked the right systems to build, and built them first. Our job is to make companies that are building a better future the winners.

How it works

Three phases: map, build, iterate.

No six-month strategy phase. No 200-slide deck. One working system beats ten pilots, and every phase ends with something concrete in your hands.

low ROItoo complexoutsource thisnot yetIdeas, ranked1.2.3.4.Q1 · liveQ2Q3
Fig. 1 · Today: everyone talking about AI, nobody steeringFig. 2 · Map: every idea captured, ranked, and agreedFig. 3 · Build: the top idea, built at startup speedFig. 4 · Iterate: a charted course through the noise

Where most companies start

AI pilots everywhere, results nowhere.

Six experiments running. Zero showing up in the numbers. The board wants an AI answer, every vendor promises the same miracle, and your competitors are already talking up their results.

The problem isn’t ambition. Nobody has aligned on which pilot is worth backing.

MapWeeks 1 to 2

Map: know exactly what to build first, and what it’s worth.

Two weeks inside your operations. Every place AI could help, scored for value, feasibility, and speed.

Built with your senior team in the room, not handed down. Leadership leaves aligned, teams want what’s coming, and new AI ideas start surfacing from everywhere.

You get: the ranked Opportunity Map, a build-ready spec for the top pick, the benchmarks behind every score, and leadership aligned behind it.

  • 3 hours of your leaders’ time
  • Yours to keep, whoever builds it
  • Worth 10× its fee, or you don’t pay

BuildWeeks 3 onward

Build: the top opportunity, built at startup speed.

Our engineers build it in your systems, with your data, alongside your team. It arrives as a working system your people actually use, not a deck about one. You keep running the business while we do the building.

You get: your first AI system live in production, built in your stack with your data, and measured against the value the map promised.

  • Priced on outcomes, not hours. You pay when the results land.

IterateEvery quarter after

Iterate: your lead over competitors grows every quarter.

We measure what shipped, improve it, and start the next build on the map. The map itself gets updated as your business and the technology move.

Twelve months in, most clients have three systems live. Competitors starting their first pilot are a year behind on data, process, and results.

You get: a quarterly shipping rhythm: the next system ships and the last one improves.

The Opportunity Map

One chart of where AI pays back in your business.

Every idea captured, scored for value, feasibility and speed, then ranked. Open the top row: that is the depth every line on your map is worked to.

Opportunity map · worked examplep.3 of 14
  • What it is
    Drafts client proposals from your own project archive, in your template, ready for partner review.
    Where it bites today
    34 fee earners touch proposals. Around 26 hours of chargeable time goes into each one, across 600 a year, and most of it is re-finding and re-typing work the firm has already done.
    What it touches
    The SharePoint project archive from 2021 onward, CRM opportunity records, and your existing template library. No new data collection.
    Evidence
    EdenLab shipped this and sits at +13% proposal output two quarters in. Every figure in the model beside this is driven off that number, so if it turns out lower for you the value falls with it rather than being defended.
    What would kill it
    Partner trust. If the draft is not good enough to send, partners rewrite from scratch and the saving evaporates. So it ships to three partners first and we measure how much of each draft survives.
    First checkpoint
    Week 3. Drafts on 20 historic proposals, scored blind by two partners against the originals that actually went out.

    How the number is worked out

    Observed at EdenLab
    +13% output
    Your proposal volume
    600 a year
    Chargeable time in each
    26 hrs
    So, extra output
    78 proposals
    At £152/hr blended
    £308k a year
    Less build and run, yr 1
    −£95k

    Year one, after cost£213k

    Recovered capacity, not cash out. It only converts if there is work to bid for, which is why the ceiling is your pipeline and not the tool.

    Score92 / 100

    • Value38/40
    • Feasibility31/35
    • Speed23/25

A worked example in the real format. Yours is built on your operations.

It has to fit how you already work.

No two businesses run the same way, so a big-ticket delivery dropped on top of one gets worked around and quietly abandoned. Whatever we build has to slot into how your team already works, without anyone changing their process to accommodate it. That is why the map starts from your operations rather than a menu of ours.

You can’t fix what you can’t see.

Two weeks buys a bird’s eye view of your own business, drawn with your senior team in the room. Most of what it surfaces is worth doing whether or not you ever build any of it with AI, so you come out of the fortnight with ideas that move the top line either way.

Results

Recent work, with the numbers.

We give our best work and our best terms to companies building things worth building: cleaner energy, healthier food, and teams that do more with every pound.

Client confidential

Clean energy · Major landowner

The next site, found faster

Their map showed new clean energy sites were queuing behind due diligence. We built a system that reads land, planning, grid and environmental data across their estate, so every candidate site arrives pre-screened and ranked.

Due diligence that took months now starts in days

PepsiCo

PepsiCo · Consumer goods

What snackers want next

Their map showed consumer tastes shifting faster than research cycles: GLP-1s and ultra-processed warnings were changing what people buy. We built a platform that analyses hundreds of thousands of social media comments, steering reformulation of key product lines and surfacing new healthier snacking opportunities.

Hundreds of thousands of comments, distilled into reformulation priorities

EdenLab

EdenLab · Energy consultancy

More proposals out the door

EdenLab’s map pointed to proposal generation. We built a drafting system on their project archive, so the team gets more proposals out, each better tailored to the client.

Proposal generation up 13%

Apply for your blueprint session.

Thirty minutes, free. You leave with a first sketch of your opportunity map, benchmarks from companies like yours, and the two or three things you shouldn’t build.

We take three blueprint clients a month. Not for companies shopping for a chatbot vendor or a slide deck.

So we can call rather than email back and forth.

Company size
Annual revenue