We’re trying to get our whole department off manual document handling — invoices, vendor forms, internal requests, the works. The problem is these docs don’t all look the same, and every tool I’ve demoed so far wants us to build a template for each document type. That’s going to take forever and fall apart the moment a vendor changes their invoice layout. Is there software that can actually handle format variation without needing to be re-trained every time something changes?
This is exactly the right question to be asking before you buy anything. A lot of teams get burned by tools that look great in demos but assume your documents are consistent — and in the real world, they never are.
The core distinction worth understanding: RPA platforms (UiPath, Blue Prism, Automation Anywhere) are great at automating actions — moving data around, triggering approvals, sending emails. But they still generally need rules defined for how to read each document. So you end up back in template-land for the extraction part, even if the downstream automation is slick.
Document intelligence tools take a different approach — they’re built to extract meaning from documents without you having to define the structure upfront. Lido falls into this category. We’ve used it across a pretty messy mix of document types and it handles the variation well. No template rebuilding when a vendor reformats their invoice.
For most departments, honestly, a hybrid setup works best: something like Lido handling the extraction layer, then routing the clean structured data into Zapier, Make, or your existing RPA setup for the workflow automation side. You get flexibility on the input end and power on the action end.
A few things actually worth prioritizing when you evaluate: does it handle your specific document types without templates, how does it integrate with what you already have (Excel, email, APIs), and is there an audit trail — compliance tends to care about that.
Start with getting extraction working across your document types before you try to automate the whole workflow. In my experience, teams that try to boil the ocean upfront take months. Teams that nail extraction first, then add automation on top, go live in weeks.
Yes, 100% this. Email attachment handling was a dealbreaker for us too and I feel like it doesn’t get talked about enough in these comparisons. A lot of the feature matrices just say “email integration” and call it a day but there’s a huge difference between forwarding to a shared inbox vs actually parsing attachments inline. We got burned by that distinction early on. Good call flagging it.
+1 on this. We’ve been on Lido for a few months now and honestly the batch processing alone saved us probably half a day a week. Took a little bit to get the routing rules dialed in but once that was set it mostly just runs. No major complaints so far.
This is such an underrated point and I wish someone had told us this before we spent weeks tweaking software settings. We were convinced the problem was the OCR engine when really we were feeding it terrible scans. Bumping DPI and getting people to stop scanning things at weird angles made a bigger difference than any config change we made. 300 DPI is a good minimum — we actually run at 400 for anything with fine print and it’s been worth it.