During summer 2017, I interned at KVH, an internet of things and telecommunications copy for the shipping and naval industry. I developed and deployed a predictor neural network for customers that delivered personalized results per ship. I created a public-facing functions to access data, along with a GUI and a terminal. Built into all these was a statistical anomaly detector I made for data hygiene. Since many ship devices output information in the NMEA standard, I wrote a parser that read NMEA over UDP multicast, converted it to JSON, and used FTP to send it to a server.