Your team re-explains
the business
to AI
every day.
MethodWorks makes sure what your leaders decide reaches the people doing the work.
Your decisions, processes and working knowledge live in a shared AI Business Brain that every person’s Claude works from. We keep it current as the business changes, so what your company learns compounds and each new decision reaches the work that follows.
MethodWorks makes sure what your leaders decide reaches the people doing the work.
Installing it takes 90 minutes with each person.
Clients include Australian businesses in financial advice, retail, logistics, education and software, and the brain stays in each client’s own storage.
processes/client-review.md
Last approved 3 October by the practice manager
Client review
Owner: practice manager. Applies to every adviser.
- 1Book the review 6 weeks before the anniversary date.
- 2The paraplanner drafts the review pack from the last advice document and the client file.
- 3Run 2 meetings. The first gathers what has changed; the second presents the advice.Proposed in the October decision session. Waiting on your approval.
- 4File notes go in within 24 hours of each meeting.
Waiting on you3
Proposed change
Client reviews move to 2 meetings
- From
- October decision session
- Owner
- Practice manager
- Changes
- client-review.md
review-letter.md - Reaches
- 12 people
Never reads what anyone types into their own Claude
Last write-back 3 October
An example brain for a 12-person advice practice. File names, roles and dates are illustrative.
brain / processes / client-review.md
Client review
- Owner: practice manager
- Approved 3 October
- 1Book the review 6 weeks before the anniversary date.
- 2The paraplanner drafts the review pack from the last advice document and the client file.
+ 2 more steps
Read by everyone’s Claude12 people
An example brain for a 12-person advice practice. File names, roles and dates are illustrative.
Ana pastes the same client
background into ChatGPT every
Monday. Raj stopped opening
Claude in week 2, and nobody
noticed for a month.
88% of organisations now use AI in at least one business function.¹
More than 80% report no tangible effect on EBIT.²
88% of organisations now use AI in at least one business function.¹ More than 80% report no tangible effect on EBIT.²
Every AI chat starts from zero, so nothing builds up. These are 5 places where that shows up in a growing business, and what each looks like once the brain holds it.
¹ McKinsey & Company, The state of AI in 2025, November 2025.
² McKinsey & Company, The state of AI: How organizations are rewiring to capture value, March 2025.
Most businesses think the answer is another hire.
Most need the team they already have off the admin.
Most businesses think the answer is another hire. Most need the team they already have off the admin.
Ana pastes the same client
background into ChatGPT every
Monday. Raj stopped opening
Claude in week 2, and nobody
noticed for a month.
Ana and Raj work at Northwind Advice, the example practice used across this page.
88% of organisations now use AI in at least one business function.¹
More than 80% report no tangible effect on EBIT.²
88% of organisations now use AI in at least one business function.¹ More than 80% report no tangible effect on EBIT.²
Every AI chat starts from zero, so nothing builds up. These are 5 places where that shows up in a growing business, and what each looks like once the brain holds it.
¹ McKinsey & Company, The state of AI in 2025, November 2025.
² McKinsey & Company, The state of AI: How organizations are rewiring to capture value, March 2025.
Most businesses think the answer is another hire.
Most need the team they already have off the admin.
Most businesses think the answer is another hire. Most need the team they already have off the admin.
You got faster.
The business behind you didn’t.
You got faster. The business behind you didn’t.
Individual AI use compounds once the business is written down where everyone’s Claude can read it.
Everything your
people re-explain,
written down once
The install interviews each person for 90 minutes and builds one shared brain from what they know. It sits in your files, under your own accounts.
- Processes
- Policies
- Decisions
- Templates
- Skills
- Objections
- Processes
How the work actually gets done, step by step, with an owner on each.
processes/client-review.md
- Policies
Your rules on AI use and client privacy, written so the tool follows them.
policies/ai-use.md
- Templates
Letters and file notes in your house style, ready for anyone on the team.
templates/review-letter.md
- Skills
Jobs each person’s Claude runs on request or on a schedule. The install includes 1 per person, and each quarterly review adds 1 more.
skills/draft-review-pack
- People and projects
Who approves what, and where each piece of work stands.
people/practice-manager.md
- Decisions
What was decided, and who owns it by when.
decisions/october-session.md
- Objections and competitors
What clients push back on and what competitors actually did. Both files grow every month.
objections.md
We never read what
anyone types into
their own Claude.
What we do read
When we run the loop with you, we read the shared brain and the systems the work already lives in: mail, calendar, CRM, ledger and project tool. Read-only. Never a staff member’s own inbox. Staff get this boundary in writing at every install.
What needs a yes
Nothing that shapes how your team’s Claude answers changes until someone with authority approves it. The record names who approved it and when.
What we readNever readWhat we readNever readWhat changes when the brain stays current
A process document starts going out of date the day it is written. The brain gets noticeably better about 2 weeks in, once your team has caught what Claude keeps getting wrong, and the loop keeps it current from there.
It doesn’t go stale
Approved changes go into the brain as they are made, so what your team reads is the current version rather than last year’s.
The same example practice as the hero, a fortnight later. File names, roles and dates are illustrative.
Decision record
Fee template updated by finance
Approved 14 October by the practice manager
- Written into
- templates/fee-schedule.md
- Reaches
- 12 people
Decision record
Fee template updated by finance
Approved 14 October by the practice manager
- Written into
- templates/fee-schedule.md
- Reaches
- 12 people
The same answer, whoever asks
Everyone reads the same approved process, so the answer does not change with who is in the room or who is away.
The same example practice as the hero. File names, roles and dates are illustrative.
Process file
processes/client-review.md
Last approved 3 October by the practice manager
- Read by
- 12 people
- Changes when
- A decision is approved and written back
Process file
processes/client-review.md
Last approved 3 October by the practice manager
- Read by
- 12 people
- Changes when
- A decision is approved and written back
New starters land in a working business
A new starter’s Claude reads the same brain on day one: your processes, your clients, your projects and your approval limits.
The same example practice as the hero. File names, roles and dates are illustrative.
New starter
Paraplanner, day one
Set up with their own voice and their approval limits
- Reads from day one
- Processes, policies, templates and client files
- Approval limits
- Set by role
New starter
Paraplanner, day one
Set up with their own voice and their approval limits
- Reads from day one
- Processes, policies, templates and client files
- Approval limits
- Set by role
You can see it working
The report card counts the hours saved, the corrections made, the gaps closed and the people who stopped writing back.
The full card is the one in the loop section below, shown with sample data.
Quarterly report card
- Time saved
- Corrections
- Gaps closed
- People writing back
Counted from your own files and compared with the quarter before.
Quarterly report card
- Time saved
- Corrections
- Gaps closed
- People writing back
Counted from your own files and compared with the quarter before.
Ask it where things are up to
Everyone reads the same brain. Leaders and managers also connect to the systems your work runs on, so a status question is answered from the work rather than from memory.
Ask it where things are up to
Everyone reads the same brain. Leaders and managers also connect to the systems your work runs on, so a status question is answered from the work rather than from memory.
Strategy that reaches the people doing the work
Every business runs this loop, whether you run it yourselves or we run it with you.
- 1ContextThe install captures how the business actually works.
- 2PrioritiesLeadership sets where AI should create value, what gets funded, who owns each initiative and when to stop.
- 3ObserveSignals come from the brain and the systems your work lives in.
- 4DecideYou make the calls on the evidence, on a fixed rhythm.
- 5ApproveNothing changes without a person saying yes.
- 6Write backApproved decisions land in the brain as updated context, workflows, instructions and skills.
- 7MeasureAgainst a baseline written down at the start.
Then it repeats. When we run it with you, every initiative is marked fix, fund, scale or stop each quarter.
Northwind Advice, Q3 2026
Quarter to date, 1 Jul to 22 Sep. The business brain and 4 people, measured against the baseline signed on 30 June: 6 named tasks, estimated the same way both times.
- Client
- Northwind Advice
- Period
- Q3 2026, to 22 Sep
- Exported
- 22 Sep 2026
- File
- pulse-NWA-2026-Q3.json
Where each person stands
| Person | Status | Last dated entry | Hours a week |
|---|---|---|---|
| SamPrincipal | Healthy | 1 day ago | 5.5 h |
| LeeAdviser | Healthy | 4 days ago | 5.2 h |
| AnaParaplanner | Watch | Today | 3.4 h |
| RajOffshore admin | Quiet | 26 days ago | 0 h |
What the numbers suggest
Check Raj’s capture before anything else
Last dated entry 27 Aug, 26 days ago, past the 21-day quiet line.
Suggested move: book 15 minutes with Raj. The laptop is personal, so Raj makes the fix with you on the call.
A quarterly report card. The target is 5 hours saved per person, per week. Each quarter adds a card to the series.
Northwind Advice, Q3 2026
Quarter to date, 1 Jul to 22 Sep. The business brain and 4 people, measured against the baseline signed on 30 June: 6 named tasks, estimated the same way both times.
Where each person stands
| Person | StatusLast dated entry | Hours a week |
|---|---|---|
| SamPrincipal | Healthy1 day ago | 5.5 h |
| LeeAdviser | Healthy4 days ago | 5.2 h |
| AnaParaplanner | WatchToday | 3.4 h |
| RajOffshore admin | Quiet26 days ago | 0 h |
What the numbers suggest
Check Raj’s capture before anything else
Last dated entry 27 Aug, 26 days ago, past the 21-day quiet line.
Suggested move: book 15 minutes with Raj. The laptop is personal, so Raj makes the fix with you on the call.
A quarterly report card. The target is 5 hours saved per person, per week. Each quarter adds a card to the series.
AI programs stall on missing decisions. Tooling takes the blame.
AI programs stall
on missing decisions.
Tooling takes the blame.
95% of enterprise generative AI pilots show no measurable P&L impact. MIT NANDA, The GenAI Divide: State of AI in Business 2025, July 2025.
Start with the brain. Add us into the loop when the business is big enough to need it.Start with the brain.
Add us to the loop when
the business is big
enough to need it.
Every business starts with the Context Install. Each size up adds one thing to what the size below it has.
| What you buy | Under 15 peopleSelf-directed between quarterly reviews | 15 to 49 peopleWe run the loop with the owner | 50 people plusWe run the loop with the executive team |
|---|---|---|---|
Context InstallBuilds the shared brain and a Claude environment for each person.Per person by role, once Explainer | Included | Included | Included |
Context LayerKeeps the brain current, with a monthly note, a quarterly review and one new skill or scheduled task a quarter for every person.Per person, monthly Explainer | Included | Included | Included |
Decision DeploymentWe run it with the owner each month: a brief, a decision session, and a quarterly review of every initiative.Per business, 12 months Explainer | Included | ||
Decision LayerThe same loop, department by department. Signals arrive weekly, and the executive team gets a monthly pack and a quarterly board review.Per business, 12 months minimum Explainer | Included |
Included
Every install ends each quarter with a report card, person by person.
- Under 15 peopleSelf-directed between quarterly reviews
- 15 to 49 peopleWe run the loop with the owner
- 50 people plusWe run the loop with the executive team
What you buy
Context Install
Builds the shared brain and a Claude environment for each person.
Per person by role, once
- Under 15
- 15 to 49
- 50 plus
ExplainerContext Layer
Keeps the brain current, with a monthly note, a quarterly review and one new skill or scheduled task a quarter for every person.
Per person, monthly
- Under 15
- 15 to 49
- 50 plus
ExplainerDecision Deployment
We run it with the owner each month: a brief, a decision session, and a quarterly review of every initiative.
Per business, 12 months
- 15 to 49
ExplainerDecision Layer
The same loop, department by department. Signals arrive weekly, and the executive team gets a monthly pack and a quarterly board review.
Per business, 12 months minimum
- 50 plus
ExplainerIncluded
Every install ends each quarter with a report card, person by person.
Add-on, any size
Decision Discipline Workshop
One room, 5 to 15 people who can fund a decision and kill one. At 15 staff that is the owner and the seniors. At 90 it is the executive team.
Any size can take it, on its own or beside the install.
ExplainerRun it yourselves,
or run it with us.
The install is the first thing we build at every size, and after that the size of your team decides who runs it. From 50 people, we usually start with the Decision Discipline Workshop.
Under 15 people
Run it yourselves
We install your AI Business Brain and set up each person’s Claude to work from it. Your team runs it from there, and each quarter a report card shows the hours saved and where the brain still has gaps.
Past 15 people, we recommend handing us the monthly side of it, on top of the brain you already have.
15 people and up
Run it with us
After the install, we run it with you each month. We bring a Deployment Brief to a decision session, and whatever you approve is written back into the shared AI brain. Each quarter you get a pack you can hand to your board.
After the install, we run it with you each month. We bring a Deployment Brief to a decision session, and whatever you approve is written back into the brain. Each quarter you get a pack you can hand to your accountant or your board.
From 50 people, the Decision Layer runs it with your executive team, department by department.
The Decision Discipline Workshop can be added whichever way you run it.
Results from live client builds, set against where each business started
Every number here has a baseline.
Multi-store furniture retailer, Australia
Their core system had no AI connector, so we built one.
Built to be handed over. The client owns every connection outright, on their own infrastructure.
Systems reachable by AI
None before. Five after, four live and one scoped.
National education charity, Australia
A drafting agent that knows when not to publish.
Consent is enforced in code, so the model never sees a blocked item. Every write is checked at the destination, so a success message means the work actually landed.
Voice rules
Built from 901 posts across 11 years.
Board advisory software, Australia
One workflow shipped, then the roadmap funded.
A support agent went live on real customer queries first. The roadmap was funded after, argued from something already working.
Elapsed
No AI in the product to a support agent live in under 3 months.
Multi-store furniture retailer, Australia
Their core system had no AI connector, so we built one.
Built to be handed over. The client owns every connection outright, on their own infrastructure.
Systems reachable by AI
None before. Five after, four live and one scoped.
National education charity, Australia
A drafting agent that knows when not to publish.
Consent is enforced in code, so the model never sees a blocked item. Every write is checked at the destination, so a success message means the work actually landed.
Voice rules
Built from 901 posts across 11 years.
Board advisory software, Australia
One workflow shipped, then the roadmap funded.
A support agent went live on real customer queries first. The roadmap was funded after, argued from something already working.
Elapsed
No AI in the product to a support agent live in under 3 months.
ASIC’s 2024 review of 23 licensees found some had AI use running well ahead of their governance.¹
Here, nothing that shapes the answers changes without a named yes. We work inside your own files and accounts, beside your team, until everyone’s Claude reads the brain. Each quarter, the report card shows the hours saved against the target, person by person.
¹ ASIC, REP 798, Beware the gap: Governance arrangements in the face of AI innovation, 29 October 2024.
Questions owners ask first
No. The layer never exports individual chats. A daily check in each person’s own account picks up decisions and project changes, and flags anything that contradicts the brain, so it stays current. If you take a decision product, we read the shared brain and your work systems, read-only.
What you can hold us to
Every person gets their own Claude
Each person’s Claude is set up with their voice, their clients, their projects and their approval limits, so they keep the work and get through it faster.
You keep the files written for you
The context files, skills, configuration and documentation written for your business sit in your own storage, under your control. Once the install is paid in full they’re yours to use in the business, whether or not we’re still in the room. We keep the rights to the method and templates they’re built from.
One thing at a time, in order
The install is the first build, then the brain is kept current, then decisions get run on it. Nothing starts before the thing underneath it is working.
Start with the brain, or start with the decisions.
Every business on the ladder starts with the Context Install. The workshop can come first when your leaders need to decide what gets funded.
Start with the brain
Context Install
We build the brain and set up each person’s Claude, in 90 minutes with each person. Everything else we sell runs on it.
Start with the decisions
Decision Discipline Workshop
You have more AI ideas than you can fund. One room of 5 to 15 people decides which ones get an owner and a date.
Book a call with Joel
Joel and the MethodWorks team put AI into the daily work of Australian businesses in financial advice, retail, logistics, education and software. Bring your headcount and the AI tools your team uses today, and you’ll leave the call knowing who your Context Install would cover.