case study

Helping Midglas transform their legacy data

Outdated client account information within the ERP system.

Ensuring client data accuracy is vital for insurance companies. Not only is up-to-date information needed to build accurate reports but also to meet regulatory compliance. Yarado helped Midglas, a Dutch insurance company that insures more than 800,000 Dutch risk addresses, to ‘clean’ and update a large amount of outdated client data within ANVA – an insurance ERP system.


Outdated accounts inactivated by Yara in 2.5 days


Manual entries automated with Yara robot


Hours from implementation to value


Up-to-date ERP system

Mitigated risk and compliant data

Increased employee satisfaction

Clean, high-quality data

Accurate customer insights

The challenge The challenge The challenge

the challenge

Due to vast amounts of outdated client account data (some over 20 years old), visibility and reporting within ANVA was very limited. Irrelevant data from 17,000 outdated accounts was hindering Midglas’s ability to derive reliable insights, polluting the more relevant current data. Furthermore, the legacy data resulted in low system performance, causing employee frustration.

The before The before The before The before The before

The before

Before Midglas brought Yara onboard, the process involved manually checking off all entries related to each client. This meant months of tedious, labour-intensive work, which was also highly susceptible to errors. The following five steps were part of this process:

  1. Selecting client account record
  2. Netting account if balance was not zero
  3. Manually ticking up to 20 entries to complete information in the client account
  4. Looking up client information associated with the account
  5. Updating client status to mark that account as inactive

The after The after The after The after The after The after The after


Besides updating the ERP system and transforming the data quality and accuracy, Yara is now periodically maintaining the system, ensuring the data quality remains at the highest level. Backed-up with advanced Optical Character Recognition and computer vision capabilities, Yara made sure that the right records were selected without missing one single client. In doing so, Yara executed two tasks:

Task #01

Netting outstanding invoices:

  1. Selecting record
    Yara selected the first client record on the list that did not have a balance of zero.
  2. Extracting entry data
    Yara opened a copy of the account information and read all entries.
  3. Applying entry data
    Entries read by Yara were ticked and afterwards the client account was saved.

Task #02

Netting outstanding invoices:

  1. Locating the client
    Yara found the client based on the unique client number.
  2. De-activating account
    After selecting the account, Yara noted down that the client data had expired and marked the last date of entry, then accepted the changes –ANVA’s account data was transformed to reflect the correct account status.
Yarado quickly updated the account data within the insurance ERP system, saving hours of manual work and transforming the quality of the data.