Case Study: Xandr achieves identity resolution at scale and personalized advertising for millions with TigerGraph

Xandr Builds an Identity Graph for Millions of Customers for Entity Resolution using TigerGraph

Xandr, the advertising and analytics division of AT&T’s WarnerMedia, needed to build Community — a digital ad platform spanning 15+ WarnerMedia properties (Cinemax, CNN, HBO, TNT, etc.) — but faced the challenge of disambiguating millions of consumers across siloed data to enable industry-first personalized commercials and advanced targeting. To solve large-scale entity resolution and identity stitching, Xandr partnered with TigerGraph and its scalable graph database technology.

TigerGraph enabled Xandr to merge siloed datasets and build a massive identity graph for people, households and devices, supporting entity resolution and cross-device advertising. The deployment runs on a 10-node TigerGraph cluster (each node: 48 cores, 400 GB RAM) and powers a distributed graph with 5+ billion vertices and 7+ billion edges, handling up to 1 billion daily updates and generating roughly 300 million vertices and 1+ billion edges from algorithms — delivering cleaner household-level data, better ML performance, more precise personalized ads and increased revenue by attracting advertisers to the platform.

Xandr

Chinmay Nerurkar

SSE2 Team Lead (Associate Director), Data Science Engineering