One person. Every connection.
Match and merge entities across Companies House, Land Registry, PSC registers, and sanctions databases using name, date of birth, and address signals.
Cross-Source Matching
The same person appearing as a PSC, property owner, company officer, and sanctions target resolved into a single unified profile across all connected registers.
Unified Profile
3 sources matched
Confidence Scoring
Match quality scored by method: exact name match, fuzzy match, DOB+address combination. Sanctions matches include confidence thresholds for risk-appropriate decision making.
Method: Name + DOB + Address
Deduplication
Matching persons identified across registers using name normalization and date-of-birth matching. Duplicate entities are automatically collapsed into canonical records.
Before
John Smith
PSC-001
SMITH, John
PSC-002
J. Smith
PSC-003
After
John Smith
Canonical record
Resolution Flow
When a person appears in multiple data sources, Glasshouse matches them using a combination of normalized name, date of birth, nationality, and registered address. The result is a single entity with all relationships from every source.
Sanctions Matching
OpenSanctions entities are matched against persons in the graph using name similarity, aliases, date of birth, and nationality. Each match includes a confidence score and the specific method used.
IVAN PETROV
Matched against graph entity
Sources
Matched Fields
Agent-Powered Resolution
Glasshouse Agents use entity resolution implicitly when tracing networks and screening. The agent follows a person across data sources, building a complete picture without manual cross-referencing.
searchPerson
Find person across sources
getPersonDetails
Fetch unified profile
traceNetwork
Map connected entities
getPersonSanctions
Screen against sanctions
Frequently asked questions.
Ready to start? Start building with a free account. Speak to an expert for your Pro or Enterprise needs.