MetaServer > Demos

Automatically classify your document by using Azure AI Document Intelligence.

Start by training a custom classification model in the Azure AI Document Intelligence Studio with at least 5 document samples. Just drag, drop, tag.

Using the trained classification model, you can feed MetaServer a single PDF containing a set of mixed document types, in no particular order. It instantly separates and identifies every document within that set. No more manual sorting or using separator sheets required.

Any low-confidence documents are indicated as orange, visually flagging the classification for quality control.

Since these flagged files are fed back into your classification model's training set, your classification model continually learns and improves over time.

With classification complete, we move to the next step: Intelligent Data Extraction.

We’ve evolved beyond the limitations of traditional OCR into the era of IDP (Intelligent Document Processing), where MetaServer focuses on interpretation on top of pure recognition.

As you can see in the demo video, whether you're dealing with delivery notes scanned on a high-end office unit or a "Mickey Mouse" mobile scanner in the cab of a truck, MetaServer isn't intimidated by poor quality.

If a human can read it, MetaServer can read it.

Another great example of the high quality data extraction using MetaServer with the power of AI.

In this demo video, we show attorney mortgage redemption letters. These type of documents used to be a real challenge.

The core data, like file numbers, dates, and names, remains consistent, but every firm uses a different layout, meaning that the data is never found in the same place on the document. This is called floating data.

MetaServer can handle this with ease by allowing you to train custom extraction models through tagging. Just click and select the required data on a few samples to train your model.

Whether it’s printed text, handwriting, barcodes, or complex tables spanning multiple pages, MetaServer can read and extract it all.

Once your extraction model has been built, you can deploy it in your MetaServer workflow.

With MetaServer, unstructured documents can return structured data.

MetaServer interprets handwritten text with human-like intuition, from messy doctor’s prescriptions to diversely written attendance lists.

If a human can read it, MetaServer can read it.

No specific training is required. Using the Azure AI Document Intelligence engine's Read or Prebuilt models, MetaServer can handle cursive, mixed formats, and multiple languages right out of the box.

In the demo video we show an attendance list: 11 different names, each written by a different person in a unique style. MetaServer effortlessly decodes each name, resulting in perfect, digital data. 

To ensure total accuracy, the system can utilize the confidence level to flag any questionable entries for validation.

Tedious, manual data entry for handwritten information becomes a thing of the past.

Using MetaServer's Validation client, you can easily verify, correct or complete your document’s data. 

To guarantee 100% accuracy, the system performs advanced checks, including database look-ups and customizable mask and format checks.

This ensures that operators can easily and reliably validate every text, date or time, line items, and amounts are before the document is further processed by your workflow.

Also, unlike traditional LLMs and AI engines that often lose visual context, MetaServer shows a side-by-side comparison of the extracted data and its exact location on the original document.

This makes verifying any data absolute child's play.

While standard PDFs only allow you to search for printed text, MetaServer can convert your image PDF to a searchable PDF. Enhanced by Azure AI, this includes handwriting, barcodes, and even stamps.

Whether it's a blurred student name, handwritten notes, or faded stamps on a deteriorated page as seen in the demo video, MetaServer reads it all.

By transforming static images into fully searchable PDFs, you can easily search for or within a document using its every single word.

While storage space is no longer a problem, ensuring a fast network connection by optimizing bandwidth remains the bottleneck for modern cloud-based workflows.

With the MetaServer Convert to PDF MRC module (= Mixed Raster Compression), you can super-compress color PDFs by a factor of 10 without sacrificing image quality.

As shown in the side-by-side comparison in the demo, even the most delicate elements like text, stamps, and highlights remain crisp and fully readable.

This technology preserves every detail of the original appearance while generating industry-standard, compliant PDFs, ensuring fast transmission and seamless integration into any system.

With the MetaServer Digital Imprinter module, you can transform document traceability by digitally embedding metadata such as processing dates, operator names, or dynamic QR codes, directly onto your documents.

With flexible configuration, you can apply these imprints to all or any specific pages, or a custom header sheet, ensuring essential information is permanently attached to the document itself.

With the built-in preview you can first verify your layout configuration during setup before applying it to your workflow.