Match & Resolve
The same person shows up in five systems under four spellings. Tilores draws edges between records using deterministic rules and fuzzy matching, then assembles them into one resolved entity — and shows you exactly why.
Toggle the rules. Watch the entity form.
Each rule draws a different kind of edge between records. Turn rules on and off to see which records merge into one entity — and how corroborating matches raise the entity score.
Real-world data doesn’t match exactly
Nicknames, transliterations, abbreviations, and typos all describe the same person. A similarity score plus a threshold you control decides what counts as a match.
Deterministic rules
Transparent, human-readable rules compare attributes — exact email, shared phone, normalized address. Every edge is tagged with the rule that made it (R1EXACT), so a resolution is fully explainable and auditable.
Probabilistic & fuzzy
For the messy middle, fuzzy algorithms and probabilistic ML score partial matches — nicknames, typos, transliterations. Choose the balance of precision and recall that fits your use case; the two approaches work together on the same entity.
See it resolve your data
Book a demo and we’ll run entity resolution against a sample of your real records — duplicates, edges, and all.