$300M combined revenue · 6 companies · 6 back offices · 12% EBITDA · NetSuite and Xero
The short version
$10.01M
More EBITDA a year
net of software and running costs
12% → 15.3%
EBITDA margin
+3.3 points
+$100M
Exit value at 10x EBITDA
illustrative, on the EBITDA added
24 weeks
All six running
first company live in week 10
01 · Where the time goes
260 people on repeatable work across six back offices. 110 after.
Full-time equivalents spent on work that is the same every week, by function, summed across all six companies.
TodayHeadless
Finance and reporting
70 → 30
Payables and receivables
80 → 32
Customer service
74 → 33
People operations
36 → 15
Total150 FTE of time back. 270,000 hours a year.260 → 110
02 · The stack
Your stack. Built by us. Owned by you.
PADISO is not in your stack. We choose the agents, wire them into the systems you already run, and set the rules inside your own tools. Every account, contract and log is in your name.
You
Operating partner and CEOs
7am portfolio summaryApprovals in SlackMonthly portfolio review
Agents
Bought, not built
DatarailsStampliUpflowIntercom FinZendesk AI agentsRipplingCustom Claude agents on n8n
The rules
Set in your own tools
Approval thresholds in SlackSpend caps in each vendorAudit logs you keepWeekly accuracy checks
Models: Claude, GPT, Gemini or Grok, chosen per task on accuracy and cost. Swappable without a rebuild.Custom agents: only where no product exists. Here, two: close prep and policy answers. Built once, deployed six times.
03 · The playbook
Function by function, step by step.
One playbook, built once and rolled out six times. The tools, the order of work, the guardrail and what it is worth across the portfolio.
Finance and reporting
All six, 3 waves
Month-end close
14 days → 4 days
Hours back a year
72,000
Net EBITDA a year
$2.67M
Tools
DatarailsNetSuiteXeroClaude agent on n8n
Steps
1Map six charts of accounts to one portfolio chart. One close calendar for all six.
2Datarails consolidates the six ledgers and builds the monthly portfolio pack.
3A Claude close agent on n8n preps accruals, reconciliations and variance notes per company.
4Company CFOs review and sign. The portfolio pack lands on day 5.
People keep
Judgement, forecasts, lender and board reporting
Guardrail
Journals over $25K and every covenant calculation need a CFO.
Payables and receivables
All six, 3 waves
DSO, portfolio average
57 days → 45 days
Hours back a year
86,400
Net EBITDA a year
$3.20M
Tools
StampliUpflowNetSuiteXero
Steps
1One supplier invoice inbox per company into Stampli. Invoices coded, matched and routed for approval.
2Upflow runs collections for all six, with one reminder standard and each company’s own tone.
3Approval limits set by company and amount. The same rules everywhere.
4A weekly cash report across the portfolio: collections, payables, 13-week forecast.
People keep
Supplier terms, credit decisions, cash
Guardrail
Every payment run is approved by a company finance lead.
Customer service
All six, 3 waves
First response
9 hours → 3 min
Hours back a year
73,800
Net EBITDA a year
$2.74M
Tools
Intercom FinZendesk AI agentsSlack
Steps
1Tag 12 months of tickets per company. Rank intents by volume and how repeatable they are.
2Deploy Fin or Zendesk AI agents on each company’s existing helpdesk. Policies written as guidance.
3Shadow for 2 weeks per company: the agent drafts, people send. Live once accuracy passes 90%.
4One escalation standard across the portfolio. Resolution rate reported weekly.
People keep
Escalations, key accounts, voice of customer
Guardrail
Refunds over company limits and key accounts go to a person.
People operations
All six, 3 waves
New starter set up
5 days → 1 day
Hours back a year
37,800
Net EBITDA a year
$1.40M
Tools
RipplingClaude agent on n8n
Steps
1Move six HR and payroll setups to one platform, company by company.
2Onboarding runs from signed offer: contract, accounts, devices, payroll, induction.
3A Claude agent on n8n answers policy and leave questions from each company’s handbook.
4A monthly people report per company: headcount, turnover, leave liability.
People keep
Hiring, performance, culture, leadership
Guardrail
Pay changes, terminations and contracts always need a person.
Tools named are what we would shortlist for this business today. No paid placements and no vendor affiliation. Best of breed moves fast, so the stack is reviewed every quarter.
04 · How it runs
Faster on every measure that matters.
Shorter bar is better. Portfolio averages. Same six companies, same leaders, same systems.
Month-end close71% faster
Today
14 days
Headless
4 days
Portfolio pack to the board75% faster
Today
Day 20
Headless
Day 5
First response, customer service180× faster
Today
9 hours
Headless
3 minutes
DSO, portfolio average21% shorter
Today
57 days
Headless
45 days
New starter set up80% faster
Today
5 days
Headless
1 day
Back office FTE per $10M revenue58% fewer
Today
8.7
Headless
3.7
05 · The money
$36.0M EBITDA becomes $46.01M.
Each function adds to portfolio EBITDA, summed across six companies. Figures are net of software, models and running the agents.
$36.0M
Today
+$2.67M
Finance and reporting
+$3.20M
Payables and receivables
+$2.74M
Customer service
+$1.40M
People operations
$46.01M
Headless
EBITDA margin
12%→15.3%
How the $10.01M is built
150 FTE × $89K$13.35MSoftware, models, running, 25%−$3.34MNet a year$10.01MExit value at 10x+$100M
06 · The 24 weeks
Three waves, two companies each. Value from week 10.
Company A proves the playbook, then it repeats. Each company runs in shadow mode for two weeks: agents draft, people send. Nothing goes live until accuracy holds above 90%.
123456789101112131415161718192021222324
Portfolio diagnostic
Diagnostic
Company A · $80M
Live wk 10
Company B · $60M
Live wk 12
Company C · $55M
Live wk 16
Company D · $45M
Live wk 18
Company E · $35M
Live wk 22
Company F · $25M
Live wk 24
DiagnosticShadow modeLive
Annual EBITDA run rate, by week
$10.01M by week 24
123456789101112131415161718192021222324
Milestones
Week 4Investment committee signs off the case, the stack and the guardrails
Week 10Company A live. Playbook proven. $2.67M run rate
Week 12Wave 1 complete. $4.67M
Week 18Wave 2 complete. Four of six live. $8.01M
Week 24All six live. Handover to Operate. $10.01M
07 · After
The operating partner’s Monday, headless.
One summary at 7am across six companies. Two decisions. The rest of the day goes on value creation.
Portfolio summaryMon 7:00am
6 of 6
companies closed by day 5
$5.6M
cash collected last week
2,310
tickets resolved without a person
Waiting for you · 2
Payment run $1.9M · Company A · 212 suppliersApprove
Journal $140K · Company C · stock write-downReview
Portfolio pack ready for the board · 13-week cash forecast updated
Always waits for a human yes
Payment runsevery batch
Journalsover $25K
Covenant calculationsevery one
Pay changes and terminationsevery one
Refunds over company limitsevery one
Illustrative model: 150 FTE of repeatable work at $89K fully loaded, less 25% for software, models and running the agents. 1,800 hours per FTE a year. Company split by revenue: A $80M, B $60M, C $55M, D $45M, E $35M, F $25M. Exit value at 10x EBITDA is illustrative, not a valuation. Not a client result. A Diagnostic replaces these with your numbers. Product names are trademarks of their owners.
What would your portfolio look like headless?
30 minutes with our founder. We sketch your portfolio before and after, and the shared stack, on the call.