From financial evidence to books you can trust.
Connect bank data, ledgers, receipts and invoices. ClosePilot matches the evidence, categorizes transactions, reconciles discrepancies, and prepares financial reports — with humans approving decisions that need judgment.
| Status | Bank / merchant | Ledger match | Amount |
|---|---|---|---|
| Matched | CloudNine Hosting28 Aug · INV-CLOUD-0826 | CloudNine Hosting Pvt LtdCloud Infrastructure | ₹4,999.00 |
| Amount mismatch | AWS India27 Aug · AWS-4481 | Amazon Web Services IndiaBank ₹15,000 · invoice ₹14,950 | ₹15,000.00 |
| Duplicate | STARBUCKS #482130 Aug · CARD-7712 | — Same card presentment as B003 | ₹1,280.00 |
| Matched | Orbit Labs26 Aug · RCPT-2408 | Orbit LabsSales Revenue | ₹24,500.00 |
Balance Sheet
BalancedCash Flow
DirectHow ClosePilot works
A single path from messy evidence to statements a reviewer can stand behind. Automation does the matching. People approve the judgment calls.
- 01ConnectBring evidence in
- 02UnderstandCategorize accounts
- 03ReconcileMatch and detect
- 04ReviewApprove exceptions
- 05ReportPublish the books
Connect your evidence
Bank data via Setu/CSV, ledger CSVs, receipts, invoices and documents. Source rows stay immutable — ClosePilot never rewrites what the bank or the books already said.
AI categorization
Understand the business and categorize transactions into the appropriate accounting accounts. Learned merchant rules and keyword policies fire first; unknown items stay uncategorized until an accountant confirms them.
Reconcile automatically
Match bank transactions against ledger entries and supporting documents. Detect matched items, duplicates and exceptions — including amount mismatches, missing receipts and one-to-many payroll groups.
Review what matters
Surface missing receipts, amount mismatches and ambiguous matches. Show evidence, deterministic checks and constrained AI analysis before a human approves. The model cannot post or change a match on its own.
Generate trusted books
Create balanced accounting entries and map them into P&L, Balance Sheet and Cash Flow reports. Category policy is company-scoped: change a mapping once, and every statement line recalculates from the same ledger.
Decision Record
Every material close decision is a record, not a chat. Evidence, matching checks, constrained AI, the proposed entry, and the human who approved it live in one place — ready for audit.
AWS India · August cloud usage
Acme India Pvt Ltd · Operating account · DR-2026-09-004
Evidence used
Matching checks
AI recommendation
Invoice and ledger agree at ₹14,950.00. The bank settled ₹15,000.00 against the same reference. Do not auto-post the bank amount. Recommend coding ₹14,950.00 to Cloud Infrastructure and flagging ₹50.00 as bank charges, pending accountant approval.
Proposed debit / credit entry
| Account | Debit | Credit |
|---|---|---|
| Cloud InfrastructureP&L → Cost of goods sold | ₹14,950.00 | — |
| Bank chargesP&L → Operating expenses | ₹50.00 | — |
| HDFC OperatingBalance Sheet → Bank / Cash | — | ₹15,000.00 |
| Balanced entry | ₹15,000.00 | ₹15,000.00 |
Human approval
“Invoice is the source of truth. Split ₹50.00 to Bank Charges so the cloud account stays tied to AWS-4481.”
Decision history
Turn evidence into books your board can trust.
Load the Acme India demo, run reconciliation, and walk a Decision Record from exception to approved entry — in minutes, not a week of spreadsheet archaeology.