Your Rules, Applied to Every Invoice.
Build custom automation rules with a visual editor. Define conditions based on vendor, amount, currency, document type, or tags — then assign actions like auto-tagging, vendor mapping, or custom AI instructions. Rules execute in pre-extraction and post-extraction phases for maximum flexibility.
Key AI Rules Engine Capabilities
Built-in functionality that eliminates repetitive document tasks
Pre and Post Extraction Phases
Pre-extraction rules run before AI processing (e.g., override language hints). Post-extraction rules act on structured results (e.g., tag invoices above a threshold).
8+ Comparison Operators
Match conditions using equals, not equals, contains, greater than, less than, starts with, ends with, and regex — covering any field in the extracted data.
Visual Condition Builder
Build complex multi-condition rules using a drag-and-drop interface. Combine AND/OR logic without writing code.
Auto-Tagging by Criteria
Automatically apply tags based on vendor name, invoice amount, currency, document type, or any extracted field — no manual sorting required.
Vendor Auto-Assignment
Map unknown vendor names to existing vendor records using rule-based matching. New invoices from recognized senders are tagged to the correct vendor automatically.
Custom AI Instructions
Override the default extraction prompt for specific vendors or document types. Useful for invoices with non-standard layouts that need targeted instructions.
What Rules Can Automate
AI identifies and extracts data from every supported format
Tag-Based Routing
Apply department tags (Marketing, Engineering, Operations) based on vendor name or invoice content for downstream routing.
Amount Threshold Actions
Flag invoices above a dollar threshold for manager review, or auto-approve small recurring charges.
Currency-Based Sorting
Route EUR invoices to the EU finance team and USD invoices to the US team using currency conditions.
Document Kind Filtering
Separate invoices from credit notes and receipts automatically, applying different tags or export destinations to each.
Vendor Standardization
Normalize vendor names across sources — "AWS" and "Amazon Web Services" both map to a single vendor record.
Custom Extraction Hints
Tell the AI to look for specific fields on invoices from a particular vendor — like a project code in a custom location.
How It Works
From connection to first extracted invoice in under five minutes
Open the Rules Dashboard
Navigate to AI Rules in your dashboard settings. The visual builder shows all active rules and their execution order.
Define Conditions
Use the condition builder to specify when a rule fires: vendor matches X, amount exceeds Y, currency is Z, or any combination with AND/OR logic.
Set Actions
Choose what happens when conditions match: apply tags, assign vendor, set custom AI instructions, or combine multiple actions in sequence.
Enable and Monitor
Activate the rule. It runs automatically on every new extraction. Review the audit log to verify rules fire correctly and adjust as needed.
Who Benefits Most
Designed for finance professionals and teams managing high-volume documents
AP Managers with Approval Workflows
Set amount thresholds and vendor-based routing to ensure invoices reach the right approver without manual triage.
Finance Teams with Departmental Budgets
Auto-tag invoices by department based on vendor or content keywords, streamlining cost allocation and reporting.
Accountants Managing Multiple Clients
Define per-client rules within each organization — different vendors, thresholds, and tags for each client engagement.
See AI Rules Engine in Action
Set up in under 5 minutes and let AI handle the busywork.
Frequently Asked Questions
Rules execute in the order they appear in the dashboard. Pre-extraction rules run before the AI processes the document; post-extraction rules run after structured data is available. You can reorder rules by dragging them.
Pre-extraction rules influence how the AI processes a document (e.g., language hints, custom prompts). Post-extraction rules act on the structured output (e.g., tag invoices where total > $500). Most users only need post-extraction rules.
Rules run in sequence. If Rule 1 adds tag "urgent" and Rule 2 removes it, the final state reflects Rule 2. We recommend reviewing rule order when adding rules that affect the same fields.
Negligibly. Rule evaluation happens in-memory after extraction and adds milliseconds to the pipeline. Custom AI instructions in pre-extraction rules may add a few seconds to the AI call.
Yes. Use the "Dry Run" feature to apply a rule against your existing invoices without changing any data. The results show which invoices would have matched and what actions would have been applied.
There is no hard limit. However, for maintainability we recommend keeping active rules under 50 per organization. Complex rule chains are better expressed as fewer rules with combined conditions.
Yes. Every rule execution is logged with the document ID, matched conditions, and actions applied. The log is accessible from the AI Rules dashboard and retained for 90 days.
You Might Also Need
Complementary tools that extend this capability