AI Financial Analysis That Shows Its Work

What the AI actually writes

The output reads like the two-page memo a good analyst would attach to your monthly pack. It opens with the headline: what happened to revenue, margin and cash this period, in numbers. Then it works through the statements: gross margin moved from 58 to 54 percent and the driver was hosting costs, operating expenses grew 9 percent against 4 percent revenue growth, receivables stretched from 38 to 51 days. Every sentence points at a figure you can find on the statement, and nothing in the memo is decoration: no filler about macroeconomic headwinds, no generic advice about watching costs, just what your numbers did and where.

The commentary sits alongside statements built by the same platform, so the income statement generator, balance sheet and cash flow output all feed one analysis rather than three disconnected reports. Founders paste it into investor updates, owners use it as the agenda for the monthly bookkeeper call, and firms ship it as the first-draft narrative in client packs.

The ratios, and what each one warns about

The analysis computes the working set of ratios that lenders, investors and experienced operators check first. Gross and net margin, and their direction over the trailing periods, because level matters less than slope. Monthly burn and runway for anyone not yet profitable, stated bluntly: at the current burn of $62,000 a month, cash lasts 7.4 months. Current ratio for near-term solvency, and days sales outstanding for whether collections are quietly financing your customers instead of the other way around.

None of these require intelligence, they require arithmetic done consistently, which is exactly the point: in AIStatements the figures are deterministic. The same upload produces the same ratios every time, computed by code, and the AI's only job is the narrative around them. The model cannot round runway up or hallucinate a margin, because it never calculates anything, it explains numbers it was handed.

Trend detection and cross-statement red flags

Single-period numbers hide problems that trends expose. The analysis tracks each line across periods and flags the patterns worth a human's attention: a margin that has declined three consecutive quarters, an expense line compounding faster than revenue, seasonality being mistaken for growth, a spike that traces back to one transaction. Thresholds are conservative on purpose: the analysis would rather surface eight items where two matter than stay quiet about the one that did.

The sharper checks run across statements, where most trouble hides. Net income rising while operating cash flow falls usually means revenue is piling into receivables. Retained earnings that do not reconcile to net income mean the books have a hole. Ending cash on the cash flow statement disagreeing with the balance sheet means an export is incomplete. These are the checks a reviewer performs in the first ten minutes of financial statement analysis, and the AI runs all of them on every upload.

Deterministic figures, AI narrative, human judgment

The division of labor is strict by design. Code parses your data, builds the statements, computes every ratio and runs every cross-check. The AI writes prose about those computed results, and each claim in the narrative is anchored to a figure the software calculated. If the data cannot support a conclusion, the analysis says so instead of guessing. That split is what makes the output usable in front of a board or a bank: you can defend every number because software, not a language model, produced it. Analysis is included on every plan, from Starter at $39 per month to Firm at $289 per month for multi-client work.

One honest boundary: AIStatements is software that formats and analyzes your data. It is not accounting, audit, tax or investment advice, it is not a CPA, and it does not guarantee GAAP or IFRS compliance. Have a licensed professional review anything you file.

Common questions