Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.dualentry.com/llms.txt

Use this file to discover all available pages before exploring further.

AI Features Overview

DualEntry embeds AI directly into the workflows you already use - posting entries, reconciling accounts, reviewing documents, and closing periods. Every AI feature operates on your live ledger data, respects your role-based permissions, and writes to the same audit trail as manual actions. Nothing happens in the background that you cannot inspect, reverse, or override.

What Is Included

Six AI capabilities ship with every DualEntry subscription. Each one targets a specific piece of the accounting workflow where pattern recognition or natural-language understanding removes repetitive effort. AI Accounting Copilot - A conversational assistant you open with Cmd+K (Mac) or Ctrl+K (Windows). You ask questions about your ledger in plain language, draft journal entries from a text description, or request explanations for flux variances during close. Copilot grounds every answer in your GL data and logs each interaction for compliance. Anomaly Detection - A nightly scan across posted transactions and account balances that flags duplicates, round-number entries, off-hours postings, and balance swings that exceed historical norms. You configure sensitivity per entity and choose how alerts reach you - dashboard widget, email digest, or close-management checklist items. Automated Categorization - When transactions arrive from a bank feed, OCR upload, or bulk import, DualEntry suggests GL accounts and dimensions based on rules you define and patterns it learns from your corrections. High-confidence suggestions apply automatically; lower-confidence suggestions land in a review queue for your approval. Bulk Update and Reclass - Select multiple records in any transaction list view, describe the change in natural language, and DualEntry parses your instruction into a preview grid showing current and proposed values. You approve row by row or all at once, and the operation records as a single reversible batch in the audit trail. OCR Document Upload - Drop PDFs or images of bills, receipts, and invoices into the Inbox. OCR extracts vendor name, invoice number, dates, line items, and totals, then populates a draft bill or invoice for your review. Corrections you make improve extraction accuracy over time. Spotlight Search - A keyboard-first search overlay that indexes transactions, master data, reports, settings pages, and recent copilot conversations. Results appear within 200 ms, grouped by type, and filtered by your permissions. Type filters like bill: or vendor: narrow results instantly.

How AI Features Interact

These capabilities are designed to complement each other rather than operate in isolation. Automated categorization and anomaly detection both feed into Close Management - categorized transactions reduce the review backlog, and anomaly alerts appear as checklist items during month-end close. The copilot can reference anomaly alerts and categorization history when you ask it questions, giving you a single conversational surface across multiple AI outputs. Spotlight Search acts as the connective tissue: any record surfaced by another AI feature is instantly reachable through a keyboard shortcut. For example, if anomaly detection flags a suspicious journal entry, you can press Cmd+K, search for the entry by number, and open it directly. From there, you can ask the copilot to explain the entry or use bulk update to reclassify it - all without navigating through menus. This interconnected design means each feature becomes more useful in the context of the others, and your overall review workflow stays within a few keystrokes of any record or insight.
Last modified on May 28, 2026