PRISM Archive – LinkThat

AI in the inbox: How MEDEWO reorganizes their emails

David Babic — Tue, 20 Jan 2026 13:07:10 +0000

Emails remain one of the most important communication channels in B2B sales – yet they are also one of the biggest obstacles to efficiency. The MEDEWO Gruppe faced precisely this challenge: thousands of incoming messages per month, manual sorting, and a significant time investment. Together with LinkThat, this process was fundamentally rethought.

With the AI solution PRISM, MEDEWO now automates its entire email inbox, ensuring that every inquiry reaches the right person directly.

When email volume becomes a bottleneck

As a leading B2B mail-order company for transport, food, and sustainable packaging, the MEDEWO Gruppe operates several brands – including MEDEWO, RAUSCH, and packVerde – throughout the DACH region (Germany, Austria, and Switzerland). The sales department alone receives around 10,000 emails per month.

Before the introduction of AI, this meant manually reviewing, forwarding, and assigning them. A process that was not only time-consuming but also prone to delays and errors.

Automated email routing with AI

With LinkThat PRISM, every incoming message is automatically analyzed – from the sender and subject line to the content of attachments such as PDFs or images. Based on keywords, context, and structure, the system recognizes the email’s content and automatically assigns it to a suitable category, such as:

Complaint
Customer Service
Order

The next step involves direct forwarding to the responsible team or the right specialist – without any manual intervention.

Dynamic, flexible, and practical

A key advantage of the solution is its continuous adaptation. Changes in team structure, vacations, or sick leave are taken into account in real time. This ensures stable response times and consistently high service levels – even under changing circumstances.

From MEDEWO’s perspective, this represents a clear benefit: shorter processing times, a noticeable reduction in workload for employees, and lower process costs.

AI that grows with you

“Our goal was not just to automate email communication, but to make it intelligent,” explains Manfred Pigl, AI expert at LinkThat. “PRISM recognizes content precisely, reacts flexibly to changes, and can be quickly rolled out to other countries and brands.”

This is exactly what the project demonstrated: After a successful launch in Switzerland, the solution was implemented at several locations in Austria and Germany within a short time.

Conclusion: increased efficiency with measurable impact

The MEDEWO example shows how AI-based automation can be pragmatically integrated into existing processes – with clear benefits in day-to-day operations. Instead of managing emails manually, AI now handles the sorting. Teams can focus on what matters most: customer relationships and value creation.

Der Beitrag AI in the inbox: How MEDEWO reorganizes their emails erschien zuerst auf LinkThat.

AI over chaos: Smart email automation for optimized customer service

Bettina Zambo — Mon, 07 Apr 2025 10:24:02 +0000

Chaos in your inbox? Customer complaints, orders, support requests – companies receive hundreds of emails every day. Without an intelligent system, it’s easy to lose track. Inquiries end up in the wrong department, are processed twice, or remain unanswered for too long. This wastes time and nerves – for both customers and employees.

With AI-powered email automation, these problems are a thing of the past. Messages and attachments are automatically analyzed, categorized, and forwarded to the right place.

From chaos to clarity: AI-powered email routing
Smart attachment analysis: why content matters more than the subject line
PRISM: more than email automation
Is your email routing future-proof?

From chaos to clarity: AI-powered email routing

Traditional forwarding systems often rely on very simple and rigid rules. But what happens when a message isn’t clearly assignable or contains multiple issues? That’s where artificial intelligence comes into play:

Automatic classification: AI uses keywords, patterns, and predefined rules to identify the type of email (e.g., support request, order, complaint) and assigns it to the correct category automatically.

Prioritization: Important emails are given priority, while less urgent inquiries are queued accordingly. Depending on the content, an email can automatically be marked as high priority.

Optimized forwarding: Replies can be routed back to the original handler – even if the email has been passed around within the company.

Smart attachment analysis:
Why content often matters more than the subject line

Not every email reveals its purpose in the subject line – key information is often hidden in the attachments. With AI-driven attachment analysis, these are scanned in advance and categorized appropriately. This includes:

Automatically recognizing document types (e.g., invoices, contracts, forms) and forwarding them to the correct department..

Extracting data from attachments – for example, reading and processing customer numbers or invoice amounts.

Analyzing images to detect elements like logos, signatures, or insurance claim details.

PRISM: More than email automation

LinkThat PRISM is our intelligent email routing solution. In addition to the previously mentioned features like classification, forwarding, and attachment analysis, PRISM offers a wide range of smart functionalities:

Email copies: Automatically sends copies to relevant departments when multiple teams need to be informed.

Duplicate detection: Repeated inquiries from the same sender are identified and handled accordingly.

Auto replies: PRISM can inform senders about the current processing status to reduce unnecessary follow-ups.

Service level monitoring: Tracks response times to improve efficiency and quality.

Autonomous processing: Automatically handles standard inquiries to relieve staff.

Compliance: PRISM meets all relevant compliance standards, including GDPR and the EU AI ACT.

Highlight: automatic tagging

For even more efficient business processes, LinkThat has placed a strong focus on automatic tagging. Historical content is automatically evaluated and categorized – no manual effort required. This applies to emails, mail, chats, and even calls.
As a result, searching for specific topics becomes fast and easy. You benefit from precise, topic-based reporting and intelligent archiving across all channels.

Is your email routing future-proof?

How efficient is your current email routing? Ask yourself the following:

How many emails still have to be forwarded manually?
Are there delays or duplicate processing of inquiries?
Do you have a clear prioritization system based on urgency?
How often do emails end up in the wrong department or remain unanswered?
Are you already using automation or AI to support your workflow?

If any of these questions gave you pause, PRISM could be the perfect solution for your business. Get in touch with us to learn how PRISM can optimally support you and your teams.

Der Beitrag AI over chaos: Smart email automation for optimized customer service erschien zuerst auf LinkThat.

Customer Journey – Trust in insurance customer service

Bettina Zambo — Thu, 03 Oct 2024 11:24:26 +0000

Trust and security are the cornerstones of insurance companies. Customers expect not only competent advice, but also fast and reliable support, especially in crisis situations. In the fifth part of our series “Customer Journey 2024 - Evolution of the customer journey”, we look at a specific example from insurance customer service. We show how to offer the best customer experience with the right tools.

Insurance companies are expected to support you when you need them. That’s why fast communication, good advice and efficient solutions are essential to ensure trust. In this use case, we take a look at how insurance companies can use AI-supported contact center solutions to improve their service quality and relieve their employees at the same time.

The most important tools for insurance customer service

Omni-channel-support

The right omni-channel solution enables all customer inquiries – regardless of the channel – to be processed efficiently and seamlessly. This allows you to maintain an overview and process inquiries seamlessly, even at peak times.

AI-supported systems

Chatbots and AI-supported systems can quickly resolve routine queries or provide personalized recommendations. This saves insurance companies valuable human resources that are needed for complex consultations.

Self-service-platforms

Insurance companies save a lot of time with a self-service platform. Customers can access all contracts, submit claims and adjust their data themselves – around the clock. This significantly reduces the workload for employees and increases customer satisfaction.

Cybersecurity and data protection

The security of customer data is our top priority. State-of-the-art security protocols protect the information and create trust, which is essential for long-term customer loyalty.

Innovative service in the insurance industry

In the insurance industry, a flexible and efficient contact center is essential. In this case study, we look at how one company is creating an evolution of the customer journey by implementing advanced solutions from LinkThat.

The initial situation

Ms. Marchetti, team leader of the customer service department of an insurance company, receives a weather warning for heavy hail the next day. To be prepared for the expected high volume of inquiries, she uses the workforce management tool from LinkThat CORE. She gets an overview of the next day’s employee resources and changes shifts to prevent staff shortages. She also informs the entire team so that all employees are optimally prepared for the upcoming damage reports.

All in one - CUBE in use

From phone, email, chat to social media – the next day, as expected, the hailstorm results in numerous inquiries. Thanks to the seamless integration of LinkThat CUBE, all channels remain clearly united in the CRM interface. The CTI middleware offers a 360° customer view and supports employees with customer recognition and contact history. This means that all data is immediately available and no manual searches need to be carried out.

Assistent ECCO is listening

LinkThat ECCO, the AI for voice channels, also acts as a helpful listener for all calls. ECCO analyzes the calls in real time and automatically provides employees with the appropriate documents, links or pre-filled forms based on predefined keywords.

Ms. Marchetti had already stored all the important links and shortcuts for the keywords “hail” and “hail damage” the day before. The agents immediately receive the relevant information and documents in the softphone. This significantly speeds up customer service in this crisis situation.

ECCO thus saves agents and customers valuable time.

Automatically to the right destination:
PRISM e-mail routing & image recognition

Many customers also send pictures of their hail damage to the insurance company by e-mail. This is where LinkThat PRISM comes in:

On the one hand, email routing with AI and attachment analysis automatically assesses every incoming message – regardless of whether it is text, images or other file attachments. Based on recognized keywords, patterns and rules, PRISM assigns the emails and takes them to the right destination. This facilitates the further processing of the numerous customer inquiries.

In the next step, PRISM takes care of image recognition. The AI is specially trained to check the validity of the images. The verified hail images are automatically categorized according to the size of the damage and forwarded to the relevant insurance department.

Analysis and reporting

Thanks to the use of powerful tools, all inquiries were answered quickly and processed further by the claims department. The day ran smoothly, with no idle time or delays. This increases customer satisfaction as well as the efficiency of the team.

At the end of the day, Ms. Marchetti evaluates the team’s performance using CORE’s analysis and reporting toolset. All important insights and KPIs are displayed here so that she can decipher optimization potential and refine future deployment plans.

With the right support, any situation in insurance customer service can be handled with ease. The combination of AI-supported tools, omni-channel support and data analysis ensures efficient employees and therefore satisfied customers.

In crisis situations, the insurance industry can create trust and strengthen customer loyalty through personalized and fast service.

Are you ready to take your customer service to the next level? Let's shape the future of your industry together.

Der Beitrag Customer Journey – Trust in insurance customer service erschien zuerst auf LinkThat.

New strategies for more efficient processing of correspondence

Bettina Zambo — Thu, 23 Nov 2023 13:41:14 +0000

We are expanding our automated document processing with two new functions. These enhancements are specifically designed to simplify day-to-day work in companies with a high volume of correspondence.

PRISM has been preparing all types of customer inquiries for further processing steps for more than five years and delivers them to the right destination. The solution is enriched by two functions that further increase efficiency in document processing: the recognition of repetitions and content segmentation.

The main task of LinkThat PRISM is to make content from letters, scans, forms, emails including attachments, faxes and checks accessible more quickly. Documents are analyzed, processed and assigned to the correct responsible departments. The following features ensure that correspondence is processed more efficiently, saving even more time.

Feature 1: Recognition of duplicates

With duplicate detection, similar and repetitive content can now be automatically identified in all documents. PRISM processes this content accordingly so that it only reaches the responsible processing department once. This saves work steps and avoids redundant responses. This results in significant time savings and increased efficiency.

The feature was developed in close cooperation with an energy supplier and can now save 15% of processing time. At the end of the test phase, the system had already correctly identified 98.5% of all duplicate content. The rate will even increase during operation.

Feature 2: Splitting document content

The second new feature is the automatic splitting of document content. This function makes it possible to recognize several different topics in one request and forward them to the appropriate processing locations. In the case of e-mails, either the original text with markings or only the relevant part can be distributed to several recipients. This enables faster and more precise processing of e-mail enquiries, as both departments immediately deal with the topics relevant to them.

“The PRISM enhancements underline our commitment to meaningful innovations that make everyday work easier.”, explains Manfred Pigl, AI expert at LinkThat. “We are responding to our customers’ current challenges and helping them to make better use of their resources.”

Supplement to existing AI functions in PRISM document processing

LinkThat PRISM retains its proven functions. The recognition and classification of content in all document types, the automatic compliance with data protection guidelines of sensitive content and the intelligent routing of documents.

The new modules were also implemented using a hybrid approach that combines artificial intelligence and program logic. “This approach allows us to achieve such high success rates.”, says Florian Jung, Head of the AI department. “Our customers receive contractually guaranteed results and can rely on the output from PRISM. The high-quality data labeling by our employees ensures this quality.”

More efficient processing of correspondence: splitting and duplicate detection now available for PRISM

Interested companies can now test LinkThat PRISM and its new functions in a proof of concept and try out all the benefits directly in their working environment.

By the way: For companies that want to drive their service processes even further forward, LinkThat SHIFT also offers a solution that enables a partially or fully automated response to all customer inquiries. Whether in chat, by email or even on the phone: SHIFT responds to customer concerns automatically with in-depth knowledge.

Der Beitrag New strategies for more efficient processing of correspondence erschien zuerst auf LinkThat.

Automated Mail Traffic for Faster Social Benefits

Harald Kerschhofer — Wed, 07 Sep 2022 11:59:44 +0000

We have extended our service management tool Myrmex with new functions using Artificial Intelligence. They are already in use today at the Austrian Social Insurance. The first project to go live was the partially automated ordering of the e-card in the summer. Insured persons benefit twice: they receive requested services faster than before, while costs in the healthcare system are saved at the same time.

“Automating the steps helps employees release each order as quickly as possible,” says Ayse Özmen-Gülmez from the development team. “All the preliminary work of verifying data across different sources is now done automatically – just seconds after submitting the online form. Employees then still check the results and confirm the application.”

On average, this saves five minutes per order. Employees can now use this time for more complex tasks. Automatic verification means that new issues can now be approved almost in real time. Insured persons receive their new cards much more quickly.

Security thanks to multi-layered checks

The security and integrity of the system were top priorities during development. Multiple interfaces are controlled to make the process as secure as possible. Depending on whether or not the applicant already has a verified login, different services are used for verification.

“What may look like a sensitive issue at first glance was actually the ideal starting point for implementing the new automation options. After all, the Austrian e-card system is digitally protected in all directions and it already covered all the necessary security standards before this project,” says Peter Kugler.

“The software accesses multiple checkpoints exclusively within the social insurance network so that every order is verified. Misuse can thus be ruled out.” Of course, this is also because service employees still confirm every application.

Fully automated processes in the private sector

For similar projects at other customers, we also offer the option of fully automating processes. This eliminates the need for an additional check at the end – after the process has always been successful in a test phase. With full or partial automation with artificial intelligence, you can react flexibly to the current and future demands of customers and partners. The best thing about it: All parties involved benefit.

Here you can see Myrmex in action. More information can be found on our product page!

Der Beitrag Automated Mail Traffic for Faster Social Benefits erschien zuerst auf LinkThat.

The Right Address: Our AI-based Email Routing

Harald Kerschhofer — Tue, 16 Nov 2021 09:04:47 +0000

Always keep an overview, even when the inbox is overflowing? This can be achieved when intelligent routing correctly assesses messages completely automatically. The content analysis with artificial intelligence from link|that looks at text, images and attachments and makes decisions based on recognized keywords. Existing email systems can also be upgraded immediately.

In practice with almost 100% success rate

In addition to classic rules that take into account the sender, text and markings, among other things, the AI also takes a close look at file attachments: it independently classifies and reads PDFs, images and documents. The system determines the best possible forwarding of the email and enables optimal processing.

In practice, the system is already being used by a large Austrian energy provider. Since the start of the project in the second quarter of 2021, over 99.9% of all incoming emails have been correctly classified and forwarded. The attachments are mostly scans or photos of documents. With AI, even emails without a subject or content – such as those often used by mailrooms to forward scanned incoming mail – can be correctly assessed based on the attachments alone.

Upgrade without detours

Integration into existing email systems is very simple: Once the incoming mail data has been entered, the evaluation and classification process begins. Emails are forwarded unchanged. link|that is also happy to help with the design of routing rules

Feel free to send questions to contact@linkthat.eu and your email will definitely end up in the right place!

Der Beitrag The Right Address: Our AI-based Email Routing erschien zuerst auf LinkThat.

How Artificial Intelligence is Paving the Way for Text Recognition

Harald Kerschhofer — Wed, 01 Sep 2021 13:54:16 +0000

How text recognition works with AI and what artificial intelligence can do in document and text recognition.

In order to check and continuously improve the quality of our AI-supported document and text recognition, we have created large test vectors. These can also be used to precisely analyze the individual steps involved in preparing documents (learn more about this here).

Submissions from the smartphone camera

When customers are asked to submit photos of documents themselves, there are often surprises: Distortions, bent or crumpled paper, blur, rotations, etc.

In the following, we would like to show you some particularly adventurous shots, which we of course took ourselves for testing. For this purpose, we printed out a publicly available contract template from BVAEB (Versicherungsanstalt öffentlich Bediensteter) and digitized it again with our smartphone in a particularly creative way.

Preparation with Artificial Intelligence

In each of the images you can see on the left our cell phone photos. On the right you can see the result from our pipeline with several AI networks that have corrected the image as best as possible. This version can then be run through the actual text recognition process. And obviously with much better results than from the original photo.

Our pre-processing manages to tighten the text and get it into straight lines – a very important preparation, so that the text can be subsequently.

Even very extreme folds in the paper can be mastered.

Two or more, as well as horizontal and vertical kinks are also “ironed out”.

Two pre-processing steps are relevant here: The crumpled paper must be overcome, but also the bending of the entire sheet. As you can see, this works surprisingly well. The shadows of the paper do look more intense (as we increase the contrast). However, the text recognition can handle this well.

This is the absolute extreme case: That customers send in their papers like this is hopefully the exception. Nevertheless, the lines of the text can still be restored quite well. The font itself becomes italicized, but that is a snap for the text recognition afterwards.

Again, the curved page results in an italicized, but very readable, new image.

And here’s a second extreme case: We clearly overdid it, because this attempt to photograph a leaf is a joke.
Still we were very pleasantly surprised by what our AI conjured out of it.

We uploaded this image with very low quality, which makes the first few sentences on paper almost illegible. On the improved version including straightening and de-blurring it looks much better. Also the result of our text recognition (to be seen below the images) surprised us very positively.

Try it yourself

If you have a lot of documents to deal with or just want to see this exciting technology in action, just drop us a line. By the way, text recognition is part of LinkThat Prism.

Der Beitrag How Artificial Intelligence is Paving the Way for Text Recognition erschien zuerst auf LinkThat.

Document Recognition from Photos: 9 Challenges

Harald Kerschhofer — Mon, 19 Jul 2021 12:47:53 +0000

Automatic document recognition is an integral part of modern business processes. But what happens when documents aren’t available as cleanly scanned PDFs, but as photos — taken with a smartphone, often under poor conditions?

This is exactly where the challenges of document recognition begin — challenges that we solve every day in our projects using PRISM.

Last updated: July 3, 2025

In this article, we’ll show you the nine steps of our processing pipeline with PRISM, which allows us to optimize even difficult photo captures so that we can recognize documents automatically — and we’ll take a closer look at the biggest challenges along the way.

9 steps for successful document recognition from photo captures

Black-and-white conversion
Brightness and contrast correction
Sharpening
Correction of alignment
Removal of noise artifacts
Depth of field correction
Perspective correction
Correction of distortions caused by bending
Wrinkle correction

Black-and-white conversion
Color tones are removed to improve text recognition – unless colored markings are relevant in the project.
Brightness and contrast correction
We precisely adjust brightness and contrast so that the text stands out clearly from the background.
Sharpening
Blurry edges are selectively sharpened to increase recognition quality.
Correction of alignment
Documents captured at an angle are realigned so that text lines can be reliably recognized.
Removal of noise artifacts
Disruptive artifacts caused by camera or compression processes are reduced using AI.

The biggest document recognition challenges in detail

6. Depth of field correction

When a document is photographed from a certain angle, only part of it may be in the camera’s focus.

The example to the right is, of course, an extreme case (you’d be surprised at the photos we’ve already seen), but it illustrates the problem well:

In this shot the depth of field is very shallow, and therefore only the front or rear part could be captured in focus. The front (green) was chosen – making the rear (red) unreadable.

In the example shown, it becomes very difficult to extract any information here. However, in less exaggerated cases, our AI manages very well to correctly process only the affected part of the shot. A smooth correction with increasing sharpness filters must be applied here. The bottom line is that this shooting error can be counteracted so well.

7. Perspective correction

In a similar situation as before, the angle may be very poorly chosen, but the content is at least sharp (or sharpened by step 6).

Nevertheless, text recognition then faces a completely different hurdle with such a photo: The text converges to the right in a trapezoidal shape; the font is much larger on the left side than at the end of each line.

Again, we have trained an AI to initiate appropriate geometric countermeasures. Thus, the font lines that converge radially regain their order, and the document as a whole can be reshaped correctly based on the detected line geometry.

8. Correction of distorted areas

The challenge of point 7 has an increase: in the picture on the right you see that the paper hangs from a table edge at a certain point. As a result, a bend starts there – and aggravating it again can be the depth of field.

Targeted training of the AI on such special cases gave us surprisingly reliable results. Especially the recognition of text lines plays its big advantage here: For a well-trained AI, fixing the unusual deformation is a snap. A challenge that would still be a Sisyphean task with conventional, logical programming shows quite clearly the advantages of self-learning AI networks.

9. Wrinkling correction (Dewarping)

Some customers seem to carry their documents around in their pockets before submitting them for further processing. Photographed documents like the one pictured on the right do indeed occur. And even if they first appear to be an insurmountable obstacle for text recognition, we can reassure you: It works!

In fact, it works so well that we ourselves were quite surprised by the results of our AI.

Conclusion: Mastering document recognition from photo captures

Document recognition from photo captures is a complex field with many pitfalls. But with the right AI-powered methods, we can optimize even the most difficult captures so that we can automatically recognize documents—reliably and precisely.

If you are also facing similar document recognition challenges, get in touch with us — we’ll be happy to support you!

Der Beitrag Document Recognition from Photos: 9 Challenges erschien zuerst auf LinkThat.