What AI nails
- The first draft, in minutes
- The grunt work: formulas, mappings, exports
- Explaining what moved, and why
“It does 80 percent of the grunt work.”Independent M&A advisor
AI-native spreadsheet · Works with Claude
Spreadsheets AI can't break.
Built with AI. Approved by humans.
Try it now
Free for individuals · No AI training on your data
Bump churn to 4% on DistribSaaS. What changes for Q1?
I ran it as a dry-run, so the saved model is untouched. Here's the Q1 impact:
Layerz patchNet revenue
€428K
−3.0% from €441K
Gross margin
€176K
−7.1% from €189K
Runway
13.4 mo
−1.6 from 15 mo
Save as scenario_churn_4% and stress a downside?
Yes, and add a downside at −15% revenue
Scenario saved. The downside cuts deeper:
Layerz patchEBITDA Y1
€212K
−51.9% from €441K
Runway
8 months
−5.2 from 13.4 mo
Covenant DSCR
0.88×
breach
Covenant breaks at month 9. Want me to surface a debt restructuring option, or just flag it on the board pack?
Plot quarterly EBITDA, base vs downside
Quarterly EBITDA, both scenarios:
Layerz readBase caseDownside −15%EBITDA · €k
Covenant floor
Q1Q2Q3Q4
The downside sits under the covenant floor from Q1 onward.
Flag it. Then export both scenarios for the board.
Done. Board pack ready:
Layerz exportDistribSaaS_Q1_board.xlsx
Export Layerz · 92 KB · 2 sheets
Standard Excel, no Layerz account needed on their end.
Write a message…
Opus 4.8
Bring your agent · connect your data
You know this file
Clean in January. Broken by June.
The problem was never building the model. It is keeping it true, close after close.
Budget_2026_v1.xlsx
Budget_2026_v7_final.xlsx
Copy of Budget_2026_v9 (Adam).xlsx
Budget_2026_v12_FINAL_FINAL.xlsx
Budget_2026_v13_THIS_IS_IT.xlsx
Budget_2026_v14_FINAL (conflicted copy).xlsx
"FINAL (conflicted copy)" is not version control.
Why finance needs more than AI
Where it breaks
#REF!
- Plain additions, silently wrong
- A different answer every run
- No memory of your model between sessions
“Claude still makes dumb calculation errors. Plain additions.”CEO, services SME
What's missing
- A structure it cannot silently break
- Checks that stay green
- A history of every change
“If there is one domain where you want deterministic, it is finance.”COO & Head of Finance, scaleup
The trail behind the number
Anyone can build a dashboard. Few can defend the number.
Behind every number: its assumption, its formula, every change since. Change the assumption, everything recomputes. Nothing breaks.
SaaS Plan 2026 – 2028
P&L202620272028
Revenue1 200 0001 900 0002 800 000
MRR900 0001 500 0002 200 000
Expansion300 000400 000600 000
COGS300 000500 000600 000
Gross margin900 0001 400 0002 200 000
% of revenue75%74%79%
Operating expenses700 0001 000 0001 400 000
EBITDA235 000441 000523 000
% margin20%23%19%
How it works
1
Start from something proven
Fork a template, bring a skill, or bring your own Excel.
SaaS budgetLBOProject financeBudget vs forecast
2
Wire your books
Set the mapping rules once. Reuse them every month.
3
It reruns every close
Variance against your frozen budget, reforecast on your actuals, a clean .xlsx out.
Your existing workflow
Same deliverables. The model that survives a handover.
Monthly close, board pack, cash forecast: standard .xlsx, no Layerz account needed on the other end. New person, same process, same numbers.
Revenue
€252K
▲ +5% vs budget
EBITDA
€68K
▲ +13%
Gross margin
77%
▲ +2 pts
March close · actual vs budget · €k
| Budget | Actual | Δ% | |
|---|---|---|---|
| Revenue | 240 | 252 | +5% |
| COGS | (60) | (58) | −3% |
| Gross margin | 180 | 194 | +8% |
| Operating expenses | (120) | (126) | +5% |
| EBITDA | 60 | 68 | +13% |
| Net income | 42 | 48 | +14% |
What you hand to your agent
How it fits your stack
The spreadsheet spirit, built to hold.
| Feature | Layerz | Excel | Claude alone |
|---|---|---|---|
| Dynamic calculations (formulas, dependencies) | ✗ raw figures | ||
| Structured for AI navigation | ✗ cells only | ✗ no persistence | |
| Variables with type + timeline | ✗ | ✗ | |
| Persists context across sessions | ✓ data only | ✗ | |
| Native financial logic (dependencies, patterns) | ✗ manual | ✗ raw figures | |
| Versioning — every edit logged, auditable, revertable | ✗ file copies | ✗ | |
| Export to standard .xlsx | native | ✗ |
Where it fits
Why not just Claude?
Claude builds a model brilliantly, once. Then a value hardens, a link breaks, nothing persists. Layerz is the surface Claude works on: Claude drives, the formulas hold, you audit every change.
Why not an FP&A platform?
Layerz keeps the spreadsheet spirit: you shape the model yourself, in minutes, and change it the day the business changes. And it stays yours, exportable to .xlsx for free, forever.
The flexibility of Excel. The confidence of a system of record.
Pricing
Free for individuals. Paid when your team collaborates.
IndividualsFree
Unlimited models. Full MCP and API access. Clean .xlsx export, always.
Teams49 €/editor/month
Working together on the same model, from the second editor in a workspace.
Full pricing →
Your data stays exportable for free. Forever. Subscribed or not.
In their words
“It brings clarity to exercises that are often complex, and, in a positive way, a bit of rigidity: traceability.”
FAQ
- What is Layerz?
- A spreadsheet built for AI. Your financial model (assumptions, formulas, the links between lines) lives in Layerz as a structure, and you drive it from Claude: ask in plain language, Claude edits the model, the engine calculates. You can also type in the grid yourself. It is the same live model either way.
- Why not just use Claude on its own?
- Claude is excellent at reasoning about your numbers. On its own, it does not keep a model: each session starts from zero, and when an assumption changes, the lines it wrote as values stop following. Layerz gives Claude a model to work on, one that persists between sessions, recalculates on every change and keeps its history.
- “In terms of crunching, it's actually faster than Claude.”(M&A professional)
- Does Claude do the math?
- No. Claude writes the structure: the assumptions, the formulas, the links between lines. The Layerz engine does every calculation, deterministically, the same way every time. So every number traces back to a formula you can open, read and question. The 10 questions to ask any AI finance tool →
- Do I have to leave Excel?
- No. Start from an Excel file you already have: Layerz imports it. Export any model back to a clean .xlsx with live formulas, any time, on any plan. Excel stays your output, and the people you send it to never need Layerz.
- What do I need to get started?
- A paid AI assistant (Claude Pro or above, or a paid ChatGPT plan) and about five minutes to add the Layerz connector. On a company-managed Claude account, your admin may need to allow custom connectors first. No credit card for Layerz.
- What happens to my data?
- Layerz does no AI processing on its side and trains nothing on your data. Your assistant runs on your own account, under your own terms with your AI provider. Your model is stored by Layerz, visible only to the people you share it with.
- I'm a fractional CFO with ten clients. Really free?
- Really free, as long as you're the only editor. That's deliberate: you're exactly who Layerz is built for.
- Who owns my model?
- You do. Every model exports in one click to a standard .xlsx with live formulas that anyone can open, and to JSON in the MDK format, whose spec is public. Its conventions live in a FINANCE.md file, an open standard any agent can read. And you drive it from the assistant you choose: Claude, ChatGPT or Codex. Open by design: you stay because it is useful, not because leaving is hard.
Spreadsheet your business.
Budgets, forecasts, board packs: built with Claude, on a structure that holds.
Try it now
Free for individuals · .xlsx export always free · Works with Claude