Ark JetStream Name and Address Matching

Overview

The Ark’s JetStream is a secure cloud-based cleanroom data matching solution with unrivalled accuracy that combines customer data with any third-party data or your own disparate data sets without sharing any personal identifiable identification (PII).  When deployed into a Snowflake cloud platform, it operates in near real-time, processing up to 3 million records per minute. Combine it with The Ark’s data sets and you achieve the very same cost saving and revenue generation benefits, but with complete automation. Combine your own disparate data sets, and you can obtain a true Single Customer View.

Background

JetStream (demo) enables you to create pseudonymous match keys from raw name and address personal data (PII). The purpose of this is firstly to create a key that can be used to match one address to another address allowing for a number of problems presented with what is basically free text data. The match keys standardise the incoming names and address to recognise that, for instance, 'flat 1, 10 high street' is the same as 'apt 1 10 high street' which an exact match cannot. The algorithm also understands that 'Jonathan' may also go by the name 'John' for instance.

Secondly, although the keys are produced from PII they contain no PII in themselves and cannot be unencrypted to reveal PII. They are pseudonymous, and can be used in place of the original personal data with a much reduced risk allowing analysis of the data to a greater audience without revealing the original personal data.

JetSteam algorithms can be used to join any name or address data not just data provided by The Ark. That may include joining datasets up across an organisation or across different organisations in a 'clean room' scenario.

Keying procedure

- Insert raw name and address data into table, pii_addresses_to_key.
- Call process_pii_addresses_to_key_proc procedure to process addresses.
- Collect results from addresses_keyed. <br/>The process passes through a 'record_id','customer_id' and a source to help join the data back up once it has been processed.

Match Keys<br/>add_key<br/>Represents a delivery point, usually an address or building. This key can be used to join address level customer data to third part data such as UK households that provides information on properties in the UK and mover alerts that shows which customers properties are on the market and therefore may be about to churn as a customer.

last_name_add_key<br/>Represents the surname within the above address, this is usually used to approximate a family where individuals share a surname within a household. It also provides a 'fuzzier' match than using first names.

first_name_add_key<br/>This key represents an individual at an address using their first and last name.

It can be used to join customer data with data such as the National Deceased Register (NDR) so that deceased customers can be identified. It can also be used to join to gone-away data such as Re-Mover where customers who have moved house can be identified.

This Demo works for a limited UK postcode and sample data is included in the application.