Dante Nardo

Viasat Outage Predictions

At Viasat, I worked on a web application to predict in-flight Wi-Fi outages and degraded quality based on a flightplan. This allowed customers to plan virtual meetings in advance and also predict when they would be unable to stream Netflix.

Computational Geometry

In order to detect outages, I implemented a computational geometry library capable of merging polygons and tracking intersections between a polygon and a line. Outages occur when a plane enters a region with no coverage or when it enters a different satellite beam. The algorithm I developed using this library efficiently calculated the outages for any line drawn across a world map with many satellite beams, like the one below.

This image depicts many polygons representing satellite coverage over a map of the world.
A sample of what satellite coverage beams looks like

Technical Details

  • Web app built using Angular, HTML, and CSS
  • API built using Java and Spring Boot
  • Computation geometry library built using Java
  • Data streamed from Kafka to Dynamo DB
  • Web app deployed to AWS via Elastic Beanstalk
  • Serverless microservices deployed to AWS using S3, SQS, and Lambda