Business Objective
Our client is a leading digital media company that is transforming mobile advertising through data-driven predictive solutions.
The critical need was to identify individuals across multiple devices they use, to maintain continuity of context while serving relevant advertisements.
Challenges
- Retrieving and identifying client mappings for several connected devices in sub-second time intervals.
- The solution needs to be fault tolerant and highly available.
Solution Methodology
- Evaluated different graph databases for the client’s use cases considering factors such as query latency, big data scalability, performance tuning capability, and cost-effectiveness
- Setup graph database cluster and ingested hundreds of GBs of data
- Performed infrastructure estimation, data pre-processing, data-integrity verification, performance tuning, REST API development, and API authentication
- Deployed the graph database cluster in production using DCOS (Data CenterOperating System) with options to provide Oauth2 authentication and load balancing
Business Impact
- The client is able to bridge online and offline channels allowing for better-targeted ads. This allows them to command a premium from advertisers through the accurate serving of ads across device-location-time combinations at an individual level.
