Our most recent Reltio Community Show featured an in-depth look at Reltio Flexible Entity Resolution Networks (FERN), our AI-powered matching engine. The session, hosted by Mike Frasca, Reltio’s VP of Product Management, provided a detailed overview of how FERN is revolutionizing entity resolution and match discovery.
Highlights from the Session
1. FERN as an Augment to Rules-Based Matching
Mike clarified a common question—FERN isn’t replacing rule-based matching, it’s enhancing it. Rules are still effective for most matches, but FERN excels at identifying subtle, hard-to-detect edge cases, such as cross-language or initial-based name variations.
2. Zero-Shot Learning with Private, Secure Data
FERN uses zero-shot learning via large language models (LLMs), which require no training on your data and maintain strict tenant isolation—ensuring complete data privacy and security.
3. What’s New in FERN 2.0
Two major enhancements were introduced:
-
Enhanced Match Discovery with Vectors: FERN now vectorizes attributes like names and addresses to identify candidates for comparison, unlocking difficult matches that traditional algorithms miss.
-
FERN Scoring for Data Stewards: All potential matches are now automatically scored by FERN, enabling data stewards to sort by relevance and prioritize high-confidence matches—dramatically boosting productivity.
4. Real-World Results
In customer testing, FERN 2.0 delivered a 7% uplift in match consolidation with no configuration or tuning, and over 10% improvement in potential match discovery.
5. Immediate Next Steps
FERN is a licensed add-on. Reltio is currently working with customers to define success criteria and conduct proof-of-concept (PoC) validations. You can enable FERN per entity type and deploy it in a phased manner.
FERN sets a new standard in entity resolution—bringing AI precision, speed, and trust to match operations.
#CommunityWebinar
#Featured
#AIML
#Matching