Skip to content
Zenta
Menu

Success story.

NLP Account Statement and Pension Projection Inquiries

Many members do not understand the technical language of their account statement and pension projections, generating doubts and massive calls after each quarterly settlement.

Image for the NLP Account Statement and Pension Projection Inquiries case

We take on any challenge

At Zenta there are no obstacles we cannot overcome.

  • Challenge

    Facilitate a self-service channel where the member asks in everyday language (“How much will I have at age 65?”) and receives an immediate, graphic and friendly response, encouraging digital adoption.

  • Solution

    We implemented an NLP assistant in Dialogflow that translates member questions into SQL queries using BigQuery. We integrated a microservice in Cloud Run to generate pension projection SVG charts and deliver them via web and mobile app. Each interaction enriches the model with AutoML Natural Language to improve understanding.

Results from our management

  • 80% of queries resolved without human intervention.

  • +45% mobile app usage in settlement quarter.

  • 30% decrease in calls to the call center post-liquidation.