Data Lake for forecasting

Challenge

The aim of the project was:

  • improving the efficiency of energy consumption
  • development of models of energy consumption prediction in terms of hourly and daily
  • identifying trends and anomalies in energy consumption data
  • unification of reporting and billing data management

Solution

A data repository in the distributed processing technology was developed.

 

Data volume: 100 TB

Result

Reducing the time of preparing billing reports from several hours to several dozen seconds

Shortening the time of preparing energy consumption prediction from a few hours to several minutes