At Radisson, we have built an AI-Driven Grouping solution which help us find groups of records. The accuracy of results shot up to 65-75% using this solution vs 20-30% as compared to match rules, for non popular/low quality data we do get unexpected results and AI tends to hallucinate at times but as I said results are far better.
For your use case, Reltio's Agent Flow (Match Resolver Agent) seems quite effective to identify matches, based on data points and hierarchies.
------------------------------
Ashish Rawat
Associate Director
Radisson Hotel Groups
Bangalore
------------------------------
Original Message:
Sent: 09-23-2025 09:58
From: Joel Snipes
Subject: AI for Entity Resolution
Has anyone here experimented with using AI for entity resolution? I want to verify the name, address, and legal hierarchy of a company in a similar way to Dunn and Bradstreet.
In theory all the data needed to do this is public. There are good open source data sets like openCorporates that aggregate US data well but international coverage is not so good
I would love to hear experiences from anyone who has tried something like this.
------------------------------
Joel Snipes
Intel
Savannah GA
------------------------------