Build Your Company Brain · A Human Layer Lab programme

We map how yourorganisation runs.Then we put AI towork inside it.

One connected system of the whole organisation, built from the tools you already pay for, with AI wired through all of it: one assistant, with the full picture, for every team. Fixed fee to map, fixed fee to build, monthly to run. You own what we build.

Book 30 minutes See pricingOrganisations of 20 to 300 people
Shape
Map, build, runthree phases, three fee shapes, no hourly billing
Price
Fixed feescoped on the call, then held to for the whole build
Platform
The Human Layer Graphyour organisation as one connected, queryable system
You own
The code and the datahanded over on request, at any time
Map · Build · Run

01 The problem

AI is only as good as what it can see.

You have been told to bring AI into the organisation. You have tried. What you got is scattered: forty people, forty private chat windows, forty different answers. Useful, individually. Invisible, collectively. Nothing compounds.

The reason is simple. AI can see almost nothing of how you actually run. The real picture lives in a CRM, an HR system, a dozen spreadsheets, a group chat and four people's heads. No one place holds it.

So we build the picture first. Every client, project, role, process, conversation and document, connected in one system. Then we wire AI into all of it.

Picture first, AI second. The only order that works. And it is built for you, not rented to you: your code, your data, your asset, worth more the longer it runs.

What we build: one connected picture

02 Three questions

Three questions AI should answer in five seconds.

You have typed these. You have had these answers. Not because the AI is not good enough, but because nothing in your organisation gives it the full picture. That is the thing we build.

Where is the Meridian rollout actually up to?

I do not have access to your delivery systems, so I cannot see where Meridian is up to. If you paste in the latest notes or emails, I can summarise them for you.

So it is answered by whoever touched it last, if they happen to be free.

Which clients have not heard from us this week?

I cannot see your inbox, your calls, or your CRM, so I cannot tell you who has gone quiet. If you paste in a list of recent contacts, I can help you check it.

So nobody knows, and it depends who remembers.

What is really going to land this quarter?

Without access to your pipeline or your invoicing I cannot say what will land. If you upload the spreadsheet, I can take a look.

So it is a spreadsheet, updated on a Friday, mostly by feel.

AI can only answer from what it can see.

03 The offer

Map. Build. Run.

Three phases, three fee shapes, no hourly billing. You own what we build.

Map

01

Fixed fee · two to three weeks

We map how your organisation runs.

Our bread and butter. The same discipline Human Layer Lab applies to organisations of thousands, applied to yours: we sit with your teams, document how the work actually moves, and build the first graph of your organisation on the Human Layer platform.

You keep the specification whether or not we build. It is yours.

Build

02

Fixed fee · priced from the specification

We build the system your organisation runs on.

The connections, the foundations, the graph underneath. Start where it makes sense: from the first graph to the full operating picture, each step live and useful before the next one is bought.

Priced from the specification, so the number you agree is the number you pay.

Run

03

Monthly · keys or retainer

This is technology you own.

Two ways to run it. We hand you the keys, or we stay on retainer, hosting, monitoring and extending it as the organisation changes.

Either way the code and the data are yours, handed over on request, at any time.

What it costs, and what you own.

Phase 01 · MapFixed feeTwo to three weeks of mapping, ending in a written specification of how the organisation runs and what is worth building.The specification is yours
Phase 02 · BuildFixed feePriced from the specification, step by step, so each step is bought on its own evidence rather than an open-ended budget.The code and data are yours
Phase 03 · RunMonthlyHosting, monitoring and extending it as the organisation changes. Or take the keys and run it yourself.Handed over on request

We do not publish a build price, because an honest one does not exist before the map. Map is quoted on the call and fixed before it starts. Build is quoted from the specification and fixed before it starts. Run is a monthly fee you can end. All fees are quoted in USD, and nothing renews by default.

Book 30 minutes

A short call to work out where AI can actually take work off your teams. If it fits, Map takes two to three weeks, and you keep the specification either way.

04 The build

AI does not arrive as a big bang.
It arrives in four steps.

Step 1

Connect

We plug into the tools your work already lives in and keep them flowing into one place. Nothing changes about how anyone works today.

AI finally has something to see.

GmailJiraSheetsDriveNotionConfluenceAsanaSAP
one place

Step 2

Read

Your teams start asking instead of hunting. Anyone can ask anything about the work and get a straight answer built from everything the organisation knows.

Nobody hunts for status.

ClaudeChatGPTGemini

Where is the Meridian rollout up to?

Scope approved Tuesday. Kick-off Thursday.

Waiting on the deposit invoice, sent 9 days ago.

Step 3

Write

AI starts doing work: updating records, drafting follow-ups, logging calls, with everything flowing back out to your other systems.

Double entry stops.

Meridian · one record

  • stagekick-off booked
  • notesworkshop logged
  • invoicedeposit sent
Jira
SAP
Gmail
Sheets

Step 4

Automate

The system watches for silence and slippage, then puts a drafted response in front of the right person. Nothing sends without a human approving it.

Nothing slips quietly.

watching · Metro quiet for 9 days

Draft ready · chase the deposit

Hi Metro team, checking in on the June invoice and where the rollout landed with your ops crew...

SWaiting on Sam✓ Approve & send

Each step is priced and delivered on its own, and each one has to earn the next. Most organisations feel the difference at step two.

05 The system

What we build, in one picture.

L1 Interface

The AI your team already uses, connected to all of it.

Asking, drafting, scoring, chasing, over MCP. One assistant, with the full picture.

L2 Screen

A command centre, built for you.

A bespoke view: every live project, its stage, and what is overdue, at a glance.

L3 Your agents

Working on the graph, directly.

Chasing, watching, scoring, drafting. Around the clock, and always a person approves.

L0 Graph

One connected graph of your organisation.

Everything, and how it relates, on the Human Layer platform. Owned by you, worth more the longer it runs.

06 In use

Same questions. Now it can see.

The same AI, the same information, for everyone. It thinks, reads the graph, and acts. And nothing sends without a person approving it.

Which clients have not heard from us this week?

checking · every client · last contact · timers

Four gone quiet, three past the point where it matters:

Metro · 9 days · deposit unpaid

Harding · 6 days · proposal sent, no reply

Kowhai · 4 days · sign-off outstanding

follow-ups drafted for all three

What is really going to land this quarter?

reading · pipeline · invoices · delivery plan

Three engagements close this quarter: Meridian, Metro, Kowhai.

Meridian finishes Friday, invoice ready to go.

One risk: Metro's change request is still unsigned.

chase drafted for the Metro change request

Get the Harding proposal ready for review.

reading Thursday's workshop notes

pricing from your rate card

checking delivery capacity

Proposal assembled: scope, team, timeline.

Built from the workshop, priced from your rate card.

Flagged: your integration lead is booked until March, so the start date moved.

in your review queue

Is this one worth taking on?

scoring against your criteria

Worth a look: right size, right sector, referred by Meridian.

One flag: their timeline collides with the Metro rollout.

reasoning written back to the record

One assistant, with the full picture, for your whole team.

07 Fit

This is not for everyone.

The programme is built for organisations of roughly twenty to three hundred people: big enough to have real process, small enough that one connected picture is achievable inside a quarter.

Worth a conversation

  • You want AI working inside the organisation, not another tool on the pile.
  • Your work moves through stages, and roughly the same ones every time.
  • Someone could explain how the organisation runs on a whiteboard in twenty minutes.
  • Between roughly twenty and three hundred people.
  • Someone on your side will own it once it is live.

Probably not yet

  • ×There is no settled way of working, because the organisation is still finding its shape.
  • ×You want to replace your CRM. Keep it. We connect to it.
  • ×There is a product on the market that already does this. Go and buy it.
  • ×The systems we would need to connect to have no way in.
  • ×Nobody internally has time to be involved for a fortnight.

Built for organisations that check.

SOC 2 Type II. Singapore region infrastructure. The same platform behind every Human Layer Lab engagement.

  • Engagement-scoped accessStaff access is scoped to the engagement, never broader.
  • Least-privilege connectionsWe connect with the minimum permissions the build needs, and you can revoke them.
  • No training on your dataEver. Your graph stays yours. Full detail in our Trust Center.

08 Questions

Asked, answered.

Individually, so does everyone. The difference is what the AI can see. Today each person's AI sees only what that person pastes in, so every answer is built on a fragment and nothing anyone learns is shared. We give one AI the whole picture of the organisation, live, and give every team that same AI. Ask it anything about the work and it answers from everything, not fragments.

No. You pay a fixed fee for the map, a fixed fee for the build, and a monthly fee for us to run it. The system itself is yours: your code, your data, your asset.

You keep everything. The code lives in a repository you own, the data in a database you can take, and the specification documents how all of it works. Any competent engineering team could pick it up. That is the point of building it properly and writing it down.

If a product exists that does what you need, buy it, genuinely. Most of our clients have already bought several. The problem is not a missing tool, it is that no tool holds the picture of your organisation, so every tool's AI is working half blind. We build the picture and connect the tools to it.

No. Step one connects to what you have. Systems get retired only if and when you decide they are no longer earning their keep, and that is your call, later, from a position of visibility.

No. It replaces the re-typing, the chasing, the hunting for status, and the Friday spreadsheet. Your people spend that time on clients and work instead. Everything the system sends is drafted for a person to approve, not fired off on its own.

Access to the tools, and your people's honesty for a fortnight. The map only works if we hear how the work actually happens, including the workarounds. The people who do the work always know where the process really lives.

Your data stays yours and is used only to run your system. Access is scoped to the engagement and logged. We connect to your tools with the minimum permissions the build needs, and we never train on your data. Fuller security detail is available on request.

Map is two to three weeks. The first build step is typically live within weeks of the specification being signed off, and it is useful from the first step, not only at the end.

Hourly means neither of us knows what it costs until it is finished, and it pays us more the slower we go. The map exists so the build can be counted, priced, and held to.

The map is built on it. Your organisation lands in the Human Layer Graph as roles, tasks, tools and processes, which is what makes the specification precise enough to price a build from. The graph stays live and owned by you long after the build is finished.

It is the sweet spot, not a gate. Below twenty there is rarely enough process to be worth mapping. Well above three hundred the shape of the engagement changes, and the Workforce AI Impact Assessment is usually the better front door.

Your organisation, running on AI. Owned by you.

30 minutes to work out where AI can actually take work off your teams, and whether how you run is ready to be built. If it is, Map takes two to three weeks, and you keep the specification either way.

We use cookies to improve your experience and analyze traffic.