Classify shipping documents

Know a shipping document without opening it

Levity lets you recognize the type of any shipping document without the need for developers or a project team.

Why bother?

Companies are spending too much time manually handling documents for shipping. Be it checking ERP information or adding a "documents complete" tag in the TMS to release the shipment.

Levity can help. With our software, you can build a system that automatically recognizes the type of document in hand, based entirely on its looks and content.

The best part? Unlike conventional document processing, our algorithms don't require a lengthy or expensive setup. All they need is a few examples for each type you wish to detect. Simple.

Trusted by leading teams

No items found.

Less worrying, more shipping

Manual handling from document receipt to storage

Machines handle routine tasks automatically

Mundane work causes employee dissatisfaction

People spend their time working on high-value tasks

Lack of insights into process data

Enhanced discovery of savings potentials

Automate what you couldn't automate before

Levity starts where rule-based automation ends

Time savings

Spend less time on handling files

Fast setup

No developers, no project team

No code

Build a custom AI without programming

Rise above mundane tasks

FAQs

You have come a long a way to get here and you might have some questions.

Couldn't we just use OCR for that?

There is nothing wrong with using OCR (optical character recognition); in fact, it is built into our software to get the best possible performance. But let me extend a bit on this issue – because there are some fine nuances that OCR-only companies tend to "forget":

Remember that bank transfers used to be instructed through hand-written forms? That's where OCR originates: Interpret hand-written letters written on the exact same layout in high volumes. The only intelligent part about it is that it could decipher a doctor's handwriting which most people couldn't.

If you are receiving the same document layout (= same template) hundreds of times each week, OCR works just as well. Set up a template once and use it until the layout is changed. But there are two traps; Trap 1: Most companies don't receive enough documents of the same layout to justify the initial investment. Trap 2: Layouts change over time.

What about extraction?

Anyone who has worked with data extraction from documents knows that it is a hard problem that no company has completely solved and neither have we. But we are working on it and parts of the problem are solved.

What exists are three things:

OCR (see above): Works well technically, expensive to set up
Systems specialized on specific document types, e.g. invoices: Works well but not for all documents
Systems that are trained on large datasets: Depends on the data

Reality check: Classification of documents is a must-have requirement for extraction to work properly and most people asking for extraction don't even have that to start with, let alone extracted all value from automating processes around it.

Having said all that, we are keen to invest in this subject and if extraction is important to you, we'd love to have a chat with you!

Will this be replaced by blockchains?

Probably. But when?

PDFs are among the worst inventions for data exchange, storage, and validation which is why blockchains are regularly brought up in the context of replacing manual work around documents. And yes, they could solve all these issues because people would just trust whatever data they see on their screen.

The problem with blockchain is that most trading partners are not using this technology and it will take years for all (including public institutions) to adopt it. Until then, PDF it is and the best you can do is find good ways to deal with those files in a (semi-) automated fashion.