Company financial model

What do you want to know from your data?

One line is enough. Step 1 asks the rest in your AI app, and nothing is done until you approve the brief it writes.

In the case folder's materials/: Your data exports (CSV or spreadsheets) and any metric definitions, or a connected warehouse.

The 3 steps

1. 01

   Agree the definitions and plan

   Then you approve, before the next step starts.

2. 02

   Compute with code

3. 03

   Explain and write up

How it does this well

One analyst, every number from code. Not found and out of scope are valid answers here: it says so instead of guessing.

How the field does this well

• Detect variances, filter by materiality stated up front, explain each with the entries behind it, and assign an owner.
• Tie out totals in code, never by the model.
• Match with deterministic rules first and leave the rest unmatched for a person; never force a match.
• Draft the narrative from plan, actuals and last period's commentary; the person supplies causality.

Ground it first

Ask for, or look in materials/ for: the plan or budget, the actuals, the materiality thresholds, and last period's commentary. If there is none, say plainly that the result rests on general defaults, and name them.

What stays with the person

Causal explanation of material variances, booking or approving entries, and sign off. Prepare these; never decide them.

The check that makes it trustworthy

Totals tie out in code, every explanation cites the entries that produce it, and unmatched items are listed.

The person's own data

If the person has connected NetSuite, Google Drive / Docs / Sheets, Microsoft 365 (Excel, SharePoint, Teams), Snowflake, BigQuery (managed remote), read their data through that connector instead of asking for exports, and say which source each number or fact came from. Without one, work from materials/.

Also from people

• Use structured instruments: the same questions, planned follow ups and a four level rubric with example answers.
• Remove names, photos and other proxies for protected traits in a script before any model reads a candidate's material. Keep the record IDs the person works with (employee, ticket or case numbers) so every finding names the row they will fix.
• Score only against the written rubric with evidence quoted.
• Draft instruments and plans; never rank people.
• What stays with the person here: every screening, hiring and people decision, and defining the role's attributes.
• The check here: proxies redacted before the model reads anything about a person, and every judgment quoted against the rubric.

Where things stay

On your computer

Your files in materials/, every result in outputs/, the work itself. It runs in your AI app, on the plan you already pay for.

With Cheatcode

The steps, the gates and how far each case has come, so a case survives closing the app. See your cases.