Select Sidearea

Populate the sidearea with useful widgets. It’s simple to add images, categories, latest post, social media icon links, tag clouds, and more.

Insurance claims NLP analyzer

Following up a successful approach by IBM using its Watson service in the insurance industry, we offered a challenging, yet exciting offer to one of our clients in insurance industry in Middle East Asia to use an Insurance claims NLP analyzer for its vast amounts of claims and boost up employees performance while increasing user satisfaction.


The main struggle of an international insurance company is processing hundreds of thousands of claims per month. Handling claims requires the intuition of highly-skilled assessors, who will pore over hundreds of pages of texts, handwritten notes, blogs, and various other sources to keep up with regulation changes and make consistent decisions.

Adding to its challenge, the degree of variation in member coverage makes it difficult to add help by training new employees.

Watson’s ability to analyze structured and unstructured data, reference the right policy information and input documents, and then make insightful recommendations, was the key factor for us and our client to help employees determine whether a claim is eligible and what percentage of the claim should be paid. This way, client employees could make better decisions and get better results faster.

To help the process and make it more efficient we used Watson Explorer  and trained it with thousands of different types of claims documents to understand the interactions, rules, and processing logic that can apply to policies. Watson Explorer is a best fit for our Insurance claims NLP analyzer system.


With IBM Watson Explorer, a cognitive exploration and content analysis platform, you can hear your data talking. Explore and analyze structured, unstructured, internal, external and public content to uncover trends and patterns that improve decision-making, customer service and ROI.


This training is ongoing and every day based on the employee interactions with the system, the system learns much more and offers better results as it goes forward. It was a big leap for our client towards scaling up to the next level.


The process as it was proposed by IBM was straightforward.


  1. A customer files a claim.
  2. Employees collect relevant information about the incident.
  3. Employees input this information into their system.
  4. Watson works behind the scenes with employees to help determine
    claim eligibility and the percentage that should be paid.
  5. Employees use a tab to access Watson’s recommendation and decide how to best proceed with payment.
  6. Employees assess claims more efficiently, saving hours of time each month.


The Insurance claims NLP analyzer system is running for almost 10 months now and has been improved drastically and also helped improve the overall employees claims assessment speed to up to %30.


July 21, 2019


AI, Deep Learning, IBM Watson, NLP