One of our main purposes has always been taking care of the insured in the moments of truth, when it needs the insurance company the most and requires it to be easy to reach, operate and to have simple efficient processes, but how?
By offering to insurance companies technological tools necessary for them to offer the best service to the clients at all times:
– Close communication.
– Faster claims processing.
– Satisfaction on the end to end process.
To achieve our objective we have gained experience in technologies such as Big Data and artificial intelligence, the last one attached to Natural language processing (NLP).
¿What is NLP?
Is a subject that combines science, systems and technology such as: artificial intelligence and linguistics on the study of interactions between computers and human beings. The objective is to provide the machine with the capacity to interpret text emulating the human ability to understand and analyze language.
NLP use cases at LISA Insurtech
On LISA Insurtech we apply several techniques of NLP. We use it with Homer (our IA) and is associated with the documents provided by the claimants while issuing a case.
This are some cases of use for Homer (IA):
-Text Classification: We train a neuronal network that allows us to distinguish between different types of documents. This allows Homer to know which type of document he is analyzing.
Example: Police Report, Passport, notice of loss
-Pattern Recognition: It guarantees the search of some patterns inside texts that allow identifying critical data on documents.
Example: RUT: 12.345.678-9
-Sentiment analysis: we use it to detect the “feeling” present while describing the event, this is, the emotional status of the one that describes the event.
Example: Anger, Sadness, among others.
-Search engines: It is a search technique based on keywords. We use it specially in the health field to quotate claims depending on the illness/injury regardless of how it’s stated.
-Example: Let’s suppose that we have a knee injury registered on our databases, inside the document provided on the claim this can also be interpreted with keywords such as fracture, knee dislocation, sprain, etc.
-Chatbots: These can be implemented by two different ways:
1.With menus that provide options for a user guide experience, usually they have a description of the possible workflows they can do.
2.Creating conversational chatbots that, with help from neurological networks, can detect the “intentions” of the claimant and answer based on it like a human.
Finally this is just the tip of the iceberg of all the things we can provide to the insurance industry.
¡For this we invite you to keep knowing about us!