Click to view the latest updates on our services.  Service Status Update.

Enhaced AI Digital Mailroom
Smart processes.
Smart people.
Smart data.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Find out more. Agree

A true digital mailroom does not stop at simply uploading the days post as a digital image, sorted by team, into a cloud-based system. Real RoI is derived when your mail starts working harder for you, removing manual data entry, and automating internal workflows.

Classification of structure mail

Visual classification logic can analyse the form and structure of each document and automatically decide if a document is to be classified as (for example): 

 

Contract

Contract

Letter

Letter

 

Customer Questionnaire

Customer Questionnaire

Invoice

Invoice

Customer Form

Customer Form

 

Ideally, any post which involves the completion of a document created by yourselves would contain a barcode. This barcode would contain important information which will make classifying and extracting information later, much more accurate.

Even when a barcode is not on a structured mail item, we can teach our systems to recognise the forms and classify accordingly.

AI Mailroom Infographic

Classification of unstructured mail

We cannot classify every document’s pre-defined visual characteristics, especially if the document is generic or lacking page structure, for example an email or traditional letter. The document may well be a customer complaint or change of address which has a workflow we can route it through. Therefore, at this stage we would classify by automating the analysing of text for key words and phrases.

Optical Character Recognition (OCR) is the conversion of typed, handwritten, or printed text into machine-encoded text.

We will push any exceptions into an “exceptions” workflow, for manual intervention. Each manual intervention feeds into our software. We use AI to make the software learn from each exception, reducing the need for manual intervention over time.

Once documents are classified, SLAs can then be applied which will ensure the mail is prioritised as per your business need.

Data extraction

Data extraction is a key element of the mailroom process, as it is the start of automation and operational efficiency. Not everyone requires this service, and a more basic digital mailroom is sufficient. However, if it is, extraction and automation allow your mailroom not only work harder for you but provide real ROI and operational efficiencies.

Once classified, being able to extract information such as invoice number, client name, invoice amount, a PO number, removes a layer of manual intervention. And once extracted this information can automatically be moved into emails, databases etc to prevent manual rekeying of information. There are three main tools for unlocking this data:

Optical Character Recognition (OCR)

Optical Character Recognition (OCR) is the conversion of typed, handwritten, or printed text into machine-encoded text.

Optical Mark Recognition (OMR)

Optical Mark Recognition (OMR) is the process of capturing human-marked data from document forms e.g. surveys & tests.

Intelligent Character Recognition (ICR)

Intelligent Character Recognition (ICR) is a handwriting recognition system that allows fonts & different styles of writing to be learnt.

Automated data validation

If the extraction process flags data as “unclear”, Restore or yourselves can manually check using an easy user interface. Highlighted fields need manual intervention. The validation software is self-learning, so when manual intervention takes place the software learns from the process to limit future manual interventions. 

Cognitive capabilities transform the way mailrooms can add value, through the understanding and interpretation of unstructured data, learning from past experiences to improve current tasks, and engaging with humans to improve interaction and experience. 

Our software can also process forms and confirm correctly completed forms. For example, highlighting ones which have not followed the rules: all boxes have been ticked, or where there are marks which might or might not be a “tick”.

In the ideal scenario, we would validate against a data set/data feed. For example, matching a policy number from a letter to a current policy number within your data set. The Restore solution can also flag within a workflow where addresses do not match up, or the policy being quoted has expired etc.

Workflow and automation

The digitised images are being pushed into an internal system already within your business or a Restore EDRMS platform, like DocuWare.

We will store all images within the chosen system, including any metadata. DocuWare, for example, can then be configured to either:

  • Have workflow folders where your teams can access the appropriate folder and begin processing the days' mail.
  • Can store the day’s mail for reference, but the information has been extracted and is “talking to” your internal databases. For example, a document classified as an invoice can be matched to a PO within your Accounts Payable systems, and the invoice details extracted and added to the AP system, saving your team the manual data entry. Another example would be AI automatically updating your CRM system of an address change.
Read more:

Read and save our Digital Mailroom brochure

Digital Mailroom Brochure

Digital Mailroom Brochure

Read and save our brochure.

Learn more >

Reconfigure your Office Space

Reconfigure your office space

Restore Harrow Green

With no onsite mailroom, we can help you reconfigure your office space.

Learn more >
Blogs you might be interested in:
Sunlight is the best disinfectant (US Supreme Court judge, Louis Brandeis)

Organisations are forever debating how best to motivate and get their teams at maximum productivity. How do you make...
Second wave of COVID makes automation and remote working even more essential, as offices look set to close again

As the second wave of COVID-19 hits, and no vaccine in sight, the Government seems on the brink of announcing that...