Book a DemoFinance AI Feature
Finance AI Workspace can work with summary-level trial balances directly, or use Data Preparation AI to reduce raw ERP exports into validated summary datasets before reporting.
Problem
Finance teams often receive ERP data at the wrong level of detail. Some exports are ready-to-use trial balances, while others contain millions of journal or invoice rows that are too detailed for monthly reporting, consolidation, or variance analysis.
Why it matters
The workspace should not force every customer through a heavy preparation step. Summary trial balances can move directly to Account Mapping AI, while raw ERP exports get prepared only when the data actually needs it.
Before / After
| Before | After |
|---|---|
| Millions of journal rows | Reporting-ready summary dataset |
| Unknown source grain | Detected summary or raw upload path |
| Manual data preparation | AI-assisted profiling and aggregation |
| Reporting from uploads | Reporting from Published Datasets |
Typical results
Finance teams typically use Data Preparation AI to:
How it works
Key capabilities
Typical finance workflow
A monthly trial balance with period, entity, account, amount, currency, and key dimensions skips Data Preparation AI and continues to Account Mapping AI.
A transaction-level export with journal, voucher, or invoice rows triggers Data Preparation AI before reporting.
The system detects ERP, language, currency, dimensions, row count, source grain, and likely reporting fields.
Raw data is reduced by month, entity, account, cost center, project, and selected dimensions.
Totals, signs, fiscal periods, currencies, duplicate rows, and mapping handoff readiness are checked.
Only a published dataset feeds Monthly Close, Consolidation Lite, Variance Analysis, Estimate, and Management Pack.
Related workflows
Screenshots
Who is it for
Inputs & roadmap
Planned integrations are based on customer demand and implementation priorities.
FAQ
No. It is optional. If the upload is already a summary trial balance, the workspace can skip Data Preparation AI and continue directly to Account Mapping AI.
It is useful when a customer uploads transaction-level ERP exports, such as journal entries, voucher rows, invoice lines, or other detailed source data that should be aggregated before reporting.
The recommended architecture keeps raw files temporary in Cloud Storage and stores only metadata, lineage, validation results, mappings, and aggregated finance facts in Cloud SQL.
Related features
Practical finance experience
Designed for teams that want the workspace to accept both lightweight summary files and larger transaction-level ERP exports without turning Cloud SQL into a raw ERP staging area.
Trust
Book Demo
Use the public page for discovery. The actual workspace remains protected behind login.