AI Finance · 30-day pilot from £3,500

Close faster, report better, forecast cash weekly. Without adding headcount.

I map how your finance team really works, fix the process, then put AI to work on the repetitive parts: invoice handling, month-end, board commentary and cash. A chartered CFO designs it and signs off every output.

Discuss a pilot See pricing
≤ 5 days
Month-end close target
−50%
Board pack preparation time target
13 weeks
Rolling cash forecast, refreshed weekly

Process first. Then AI.

Automating a broken process just makes the mess faster. Every engagement starts by mapping what actually happens, so the AI goes where it saves time and the controls stay intact.

01 · Diagnose

Process and controls review

An end-to-end map of how invoices, approvals and close really run: every owner, handoff and system step. You get a clear view of bottlenecks, control gaps and what can and can't be measured today.

02 · Close & Control

Faster, cleaner month-end

Invoice intake and coding support, reconciliation and anomaly flags, accrual prompts and a close calendar. AI handles the heavy lifting on complex coding and allocations, such as multi-line invoices with mixed VAT treatment.

03 · Insight & Board

Margins, commentary and board packs

Margin analysis by service line or customer, with the levers sized. Variance commentary drafted in minutes, and a KPI and board pack template, reviewed and signed off by a CFO before it goes anywhere.

04 · Cash

Cash model and forecast

Total available cash including facility headroom, a cash bridge that explains every month, a 13-week to 15-month forecast, and the levers that move cash most, with runway and covenant flags.

Recent work

What this looks like in practice

Case study · Cash

From month-end cash anxiety to a cash model the CEO can use

A UK-headquartered, multi-country services group funded by invoice finance. Leadership felt a worsening month-end squeeze but couldn't pin down why.

What was built

One workbook measuring total available cash (bank plus undrawn facility), not just the bank balance. It includes a month-to-month cash bridge with a plain-English explanation, reconciled debtor, unbilled and creditor movements, key-customer tracking, a 15-month projection with editable assumptions, a dashboard ranking the ten levers that move cash most, and a how-to guide for the CEO.

How AI was used

AI read and reconciled around 60 source files line by line (management accounts, consolidations, aged debtors and creditors, accruals and bank reports), then built the model, ran scenarios and drafted the CEO summary. CFO challenge caught its mistakes, such as a misleading supplier-payment conclusion that vanished once accruals were included. The AI in turn caught errors in existing reports, including a stale data feed and a reversed FX sign.

~3 months
of sales tied up in unpaid invoices and unbilled work: the root cause, found
£100k → £190k
each extra £100k of monthly sales absorbed about £190k of cash
10 days > 2 pts
faster collection was worth more cash than a 2-point margin gain
~10 min
monthly refresh with AI, once month-end files are in

The model also quantified the case for a separate overseas funding facility, ready to take to the bank.

Case study · Margin

A margin improvement model for a mid-market services business

A multi-country services provider, much of it subcontracted work, running at a thin EBITDA margin under cash pressure.

What was built

A fully formula-driven model linked to the management accounts. It splits the P&L into four margin layers, analyses gross margin across nine service lines, sizes ten margin levers at conservative, target and stretch levels, and ends in an EBITDA bridge and a phased action plan with owners and KPIs.

How AI was used

AI reconciled the management accounts, budget, cash model and bank data, found the diagnostic patterns, sized each lever from actual cost lines, and built and checked the workbook: 756 formulas, zero errors, EBITDA tied to the management accounts to the pound. The judgement calls stayed with the CFO.

27% → 21%
gross margin slide in the largest service line, traced to wages up 11% as revenue fell
~10%
of revenue found in two service lines earning near-zero margin
3.5% → 12%
EBITDA margin at the target scenario, with central overhead the biggest lever
~14 months
phased path to around a 10% run-rate

Diagnostic findings are from the client's own data. Scenario and phasing figures are modelled projections, not yet realised.

Case study · Process

Accounts payable review for an international services group

Two ledger companies, UK and US, three invoice streams and around 630 supplier invoices a month, with a stretched AP team.

  • Mapped every step across four departments, from inbox to payment run, with swim lanes for each invoice stream.
  • Found 11 handoffs, only 2 controlled by the system. Delays clustered at the uncontrolled ones.
  • Showed that invoices were invisible until keyed into the ledger, so ageing, accruals and time-to-pay could not be measured. One fix closes both the reporting gap and the control gap.
  • Built a volume and value baseline from seven months of ledger data, to measure improvement against.
  • Identified where AI is already saving time, such as complex telecoms allocations, and where automation should go next.

Start with a 30-day pilot

The pilot spans a real month-end, so you see measured results on your own numbers before committing. All fees ex VAT.

30-day pilot

Start here
From £3,500
  • Week 1: process diagnostic and baseline
  • Week 2: commentary and board pack workflow
  • Week 3: 13-week cash forecast
  • Week 4: run live through month-end, measure results
  • Fee credited against setup if you proceed
Discuss a pilot

Full setup

From £8,000
  • All four modules, built around your systems
  • Process fixes and control improvements
  • Playbooks and templates your team keeps
  • Training for your finance team
  • Priced by entities, data sources and pack complexity
Get a quote

Ongoing retainer

From £1,500/mo
  • Core: monthly commentary and board pack, cash forecast refresh, review call
  • Plus (from £3,000/mo): close oversight, weekly cash, board attendance, continuous improvement
Talk it through

Who it's for

  • UK SMEs and scale-ups, roughly £2–30M revenue
  • Small finance teams with no FP&A capacity
  • On Xero or Microsoft Dynamics (NAV / Business Central)
  • Boards wanting better monthly reporting
  • Close taking more than 10 days, or a key finance person leaving

Safe by design

  • AI drafts; a chartered accountant reviews and signs off
  • Every number reconciles back to your ledger. The AI is never the source of a figure
  • Read-only access to your accounting system
  • Business-grade AI tools, with data use covered in the engagement letter
  • Approvals and segregation of duties stay intact

Questions

Will this replace my finance team?

No. It takes the repetitive work off them so they can focus on review, exceptions and decisions. Most clients use it to absorb growth without hiring.

Which systems do you work with?

Xero and Microsoft Dynamics (NAV / Business Central) today, alongside tools like ServiceNow and Excel. If you use something else, ask.

Is our financial data safe?

Access is read-only, tools are business-grade with data protections, and AI processing is set out in the engagement letter and data processing terms before anything is connected.

What if our books aren't in good shape?

That's what the diagnostic is for. We fix the basics first, then automate. Sometimes the biggest win is the process change, not the AI.

See it on your own numbers

Start a pilot before your next month-end and measure the difference.

Discuss a pilot