ع
Start Topics Teams Reference What's new Saved
tracks

Finance & Ops

Wrangle spreadsheets, reconcile exports, run the recurring ops summary, and turn raw numbers into clear answers — with the formulas explained — all from the Claude Desktop chat, no terminal required.

11 playbooks
playbooks

Full worked systems, not one-off prompts — each walks the whole workflow end to end. Pick one and follow it.

Make a messy export trustworthy

Profile a raw export before you trust a single total — row count, what each column means, blank rows, duplicate IDs, mismatched dates — and walk away with a clean version you can build on.

easy ~20 min open the playbook

Build the categorization rulebook every report inherits

Turn your chart of accounts and the way you actually categorize spend into one categorization-rules.md Claude applies every time — so the same vendor lands in the same bucket across every audit, budget, and close, instead of being re-guessed each month.

medium ~1 hour to build, reused every month open the playbook

Reconcile two files, line by line

Match invoices against payments (or bank against ledger) by ID, list everything unpaid or unmatched, total the outstanding, and flag any payment with no invoice — a monthly check you can defend line by line.

medium ~40 min open the playbook

Audit the month's spend

Scan transactions for anything unusual — charges far above normal, duplicates, vendors you've never paid — suggest a category for every uncategorized row, and surface the one line that needs explaining to leadership.

medium ~40 min open the playbook

Track budget vs. actuals and explain every variance

Line up the plan against what actually happened, flag every variance over your threshold, explain each one as price, volume, or timing — not just a number — and write the variance narrative leadership can act on, with the math kept where you can re-run it.

medium ~45 min open the playbook

Consolidate vendor & SaaS spend and catch renewals

Pull every vendor and subscription into one view, catch the duplicate tools, the silent price hikes, and the seats nobody uses, and build a renewals calendar — so nothing auto-renews at 3x by surprise and you can cut spend with evidence, not vibes.

medium ~45 min open the playbook

Run AR aging and chase collections

Age your receivables into 0–30 / 31–60 / 61–90 / 90+ buckets, rank who to chase by amount and overdue days, draft the right reminder for each bucket, and track DSO month over month — so cash you've already earned actually lands instead of aging quietly.

medium ~40 min open the playbook

Build a cash-flow model you understand

From a starting balance and recurring costs, project the next 6 months with every formula explained in one plain sentence — so you stay in control instead of trusting a black box.

medium ~1 hour open the playbook

Brief leadership in plain English

Turn a month of numbers into a narrative leadership will actually read — what changed, why, and the two things to watch, under 200 words — while keeping the formulas and raw figures where you can re-run them.

medium ~30 min open the playbook

Run month-end close, end to end

Clean the export, reconcile the accounts, roll up totals with shown work, compare to last month, model the runway, and write the leadership summary — a repeatable close where every number is verified and defensible.

advanced ~half a day open the playbook

Assemble the board & investor finance pack

Pull the close, the model, the KPIs, and the budget variance into one board-ready pack — the metrics investors track, a CFO-grade narrative, and an appendix where every number traces back to a verified source — so you walk into the meeting able to defend any line on the page.

advanced ~half a day open the playbook
Go deeper

Finance & Ops with Claude, in depth

The playbooks above are single jobs. These are the long reads behind them — how the work fits together across a quarter, what good looks like, and the judgement calls the prompts assume you have already made. Read in any order.

Guardrails

The non-negotiables before any of this ships. Read the operating guide for the full data, brand, legal, and rollout playbook.

  • Ask Claude to check the file for problems first — blank rows, duplicates, mismatched date formats — before you trust any total.
  • Always ask it to show the formula or steps it used. Finance math should never be a black box you can't verify.
  • Spot-check one number by hand against the raw file. If it matches, you can trust the method on the rest.
  • Keep sensitive financial data in your own approved workspace, approve each read in the 'Ask permissions' prompt, and let Claude work the local copy — never paste account numbers into tools you don't control.
  • When Claude cleans an export or saves a summary, review the accept/reject diff so you see exactly what changed before it lands.
  • Have Claude write the summary in plain English for leadership, but keep the numbers and formulas where you can re-run them.
  • Keep the person who prepares a number separate from the person who approves it for anything that moves money, and trace every reported figure back to a source file — so the work is reviewable, not just done.
Read the operating guide

Questions people ask

Can I trust the numbers Claude gives me?
Only after you check them — which is the point of asking it to show its formula or steps. Spot-check one total by hand against the raw file; if it matches, you can trust the method on the rest. The Desktop visual diff also lets you see exactly what a cleanup changed.
Is it safe to put our financial data into Claude?
Keep account numbers and sensitive financials in a workspace you control, and let Claude read the local copy through the Desktop file pane, approving each read in the 'Ask permissions' prompt. Working from an export that way is fine — and always verify totals before you report them.
Does this replace my finance or ops team?
No. It removes the manual grind — cleaning exports, reconciling files, building the recurring summary — so your people spend time on judgement and review. A human still owns every number that goes out.
What's a realistic first task?
Cleaning a messy export: drop a billing or expense CSV into the Desktop chat and ask it to flag blank rows, duplicates, and mismatched dates before you total anything. Catching dirty data first is the highest-value ten minutes you'll spend.