By using Google’s cloud-based analytics and machine learning technologies, UK broadcaster Sky is giving itself the means to combat piracy of Premier League football matches. A data lake is also built to exploit its data massively.
UK broadcaster Sky has turned to the Big Query data warehouse and Google Cloud’s automatic learning algorithms to investigate web traffic and shut down hacked sites in real time. Speaking on stage at Next 2018 in San Francisco this week, Mohamed Hammady, CTO of Sky UK said the company invests “nearly $8 billion a year in content. A major part of this investment is used to purchase match rights. Moreover, the crown jewel is the British Premier League. “Sky pays $1.6 billion a year for the rights to these 126 matches. Piracy is, therefore, an obvious threat to this investment.
After the historic High Court decision in 2017, which was strengthened earlier this month to stop server-level traffic, ISPs are now responsible for closing down illegal sites. “Unfortunately, as in any industry, some malicious people try to broadcast football matches illegally,” says Mohamed Hammady, a little naively, “this can harm the sports industry and cannot be tolerated.
500 billion data records per year
Despite everything, this court order did not resolve the technical problem of real-time identification and these sites. Sky, which is in the dual position of sports broadcaster and Internet service provider, has therefore chosen Google Cloud to resolve this.
Sky’s team collected its traffic data on NetFlow to sample traffic on the core network. For one year, 500 billion data records were produced. Sky, therefore, needed a highly scalable data warehouse solution to manage this. “Using BigQuery and an internal algorithm – which cost $10,000 to develop – we are now able to continuously study traffic patterns with an always up-to-date list of suspicious hacker sites,” says Mohamed Hammady. “Once they have been confirmed as illegal, they are closed. »
GCP chosen to build a data pool
SKY’s CTO continues: “The query execution time on Google Cloud is less than 30 seconds and costs 23 cents for each use. It is the 23 cents of my PnL [Profits and loss] that I am most proud to spend. This has led to a remarkable reduction in pirate sites in the UK”. “This experience led Sky to turn to Google Cloud for its larger data warehouse needs. “We decided to build our data lake on Google Cloud Platform,” adds Mohamed Hammady.
The critical element of this internal data strategy is to better customize its services to its customers by consolidating their data in one place. “This ranges from content recommendations to service calls, across all channels,” concludes Hammady.