Skip to main content

July 2026: Meet Capture, and a few Helpdesk wins

Release notes July 2026

This month, we're introducing Capture, our new invoice extraction module that fixes one of AP's first workflow hurdles: extracting invoices with 99% accuracy. We've also shipped a couple of smaller Helpdesk improvements that cut down on manual admin.


Capture šŸš€

Capture is the invoice intake layer of Xelix. Using LLMs, it reads, corrects and validates any invoice like a human, at 99% extraction accuracy.

Agents you build in natural language then resolve exceptions automatically, cutting the cost and manual work that invoice processing creates today.

Capture aims at resolving exceptions at intake, so more invoices reach your ERP, increasing your straight-through-processing rate.

Here's what's included:

  • Reads invoices across any layout, including handwriting

  • Configurable agents in natural language - set up exception handling and business rules by describing them in plain language, not by raising an IT ticket

  • Automatic exception resolution - agents correct fields, apply rules by line item or globally, and auto-send invoice rejection emails with the correct context

  • Human-in-the-loop by design - every agent is proven against historic data and tested before going live, and every action it takes is logged for a clean audit trail

Full launch details are covered in our Capture blog post here! You'll find out what Capture does for your AP team, how it's different from outdated OCR systems and where it fits into your workflow streams.

What this means for you:

→ Invoices extracted accurately from day one, no supplier setup required
→ Exceptions resolved automatically instead of landing on someone's desk
→ A foundation for higher straight-through processing across your AP process


Helpdesk

1. Automated Divisions

We've made Reminder and Invoice tickets automatically inherit the right division(s) whenever they're matched to an invoice with high confidence.

What this means for you:

→ Less time spent manually tagging tickets by division
→ Fewer tickets missed or mis-assigned between divisions
→ More reliable division-level reporting, without relying on manual entry


2. Kanban All Mailboxes View

We've made Kanban open on "All Mailboxes" by default, so you see tickets across every mailbox and folder in one place. Per-mailbox and folder filtering still work exactly as before, just one click away.

What this means for you:

→ Full team visibility the moment you open Kanban
→ Less time lost piecing together separate mailbox views
→ Still one click away from a focused, single-mailbox view

Did this answer your question?