Supporting the IT department of Bouygues Telecom to improve the QOS by detecting and predicting anomalies using big data.

Our work


In order to improve the quality of service, our teams aggregate, transform and restore data from all Bouygues network equipment (internet boxes, smartphones, routers, etc.).

This data is then used in two different ways: for real-time dashboards or via BI tools and machine learning algorithms (for anomaly prediction).


We were responsible for all project phases (monitoring, design, implementation, and maintenance) and made recommendations that the client validated.

As the infrastructure is on-site, we controlled our infrastructure down to the last detail and installed the management and automation solutions ourselves.

Skills and Roles

A service centre of 9 people:

  • 1 project manager
  • 4 developers (1 manager incl)
  • 4 DevOps (1 leader incl)

Technical Frameworks

Java, Spark, Tenserflow, HDFS, ElasticSearch, Kafka, Hive, Cassandra, Hortonworks, Docker, Ansible, Kubernetes.

5 Petaoctets of aggregated data
77 Nodes deployed
150k events/sec. on the real time part

Related references