There’s a lot of noise around AI and identity right now.
Hightouch just launched Adaptive Identity Resolution, Meiro is leaning hard into real-time CDP-native matching, and Amperity continues to push its probabilistic + deterministic blend for enterprise brands. But when someone asked me whether GenAI could solve identity resolution, I went straight to someone who’s seen the inside of the problem: Steven Renwick, co-founder and CEO of Tilores.
Steven and his team aren’t building another black box, they’re focused on traceability, reversibility, and real-time performance for messy, high-volume data. In this episode, we dig into what that actually means in practice.
💬 Topics we cover:
Why Tilores avoids GenAI for core matching—and where it does use LLMs
Entity graphs, not vector embeddings
Using RAG (retrieval augmented generation) to query resolved identities
Matching for fraud use cases, telecom mergers, and multi-brand insurance data
Why explainability still beats hype in regulated industries
“Fraudsters aren’t always smart. They just hope your data isn’t smart either.”
We also touch on identity matching horror stories, Cockney rhyming slang, and why test data is the silent killer of your data quality.
Learn more about Tilores → https://tilores.com/
👥 Connect with us on LinkedIn:
📲 Follow Martech Therapy on:
Share this post