- A computer distribution company in Americas running Salesforce CRM with 5.2 Million records successfully consolidates the CRM of its acquisition in Asia which had 1.0 Million contacts in its legacy system, within a month.

- A hospital in Philippines gets data in different formats from disconnected systems of labs and pharmacy. They are able to chain these clusters of data into a single record in their Patient Management System.

- An Asian manufacturer of Hearing Aid devices is able to import data from 75+ resellers in inconsistent format and structure it within its CRM and Customer Data Platform automatically.

Connecting Records for Single Version of Truth

Contactous' Enterprise Data Hygiene (EDH) is being used to find common patterns across disjointed data clusters  and stitch them together to give a 360 degree view of the record. It is best explained by an example. Assume that there are 5 datasets with an average of a million records in each, with some overlapping fields between them. EDH is able to chain these duplicate records out of the millions and create a unified view within seconds. 

In this example, EDH has created these 5 records in a cluster from 5 million records within seconds. It does this by:

1. Matching name and then name + date between 1st and 4th record. It does a double check by matching state and country too. It now has a mobile number (from 4th record).
2. Using the mobile number, EDH then chains 2nd record and gets email and ID number. 
3. Connecting 2nd and 3rd record with email address follows. 
4. EDH goes back to output of Step #1 and uses the twitter handle to connect to 5th record.
5. The result is on the right - A unified data record from 5 disconnected datasets. The record is also cleaned and standardized as shown in fields of date and address.​

An actual use case of golden record creation

The algorithms of EDH to link duplicates towards a unified view are used to chain records at a hospital for patient data. In the illustration below, EDH has found 4 records from 4 different datasets and then consolidated them into a specific one. 

FAQ :: Customer Golden Record

Can the system cluster duplicates and merge them to a unified record by itself?
Yes, you have the option of letting EDH merge the found records into a specific dataset, based on your preferences. 

You refer to an approach called 'Chain Duplicates'. What is that? 
For years we have has focused on creating algorithms to detect complex duplicate patterns. These algorithms work in specific dimensions - eg, Person's Name, Phone Numbers, URLs, Addresses, Company Names etc. Chain Duplicates is an approach that we use to unify the output of our existing DeDuplication algorithms which process data from fields of disjointed datasets. 

My disjoined datasets are in hundreds of CSV files. How can they be chained? 
EDH has a function to upload CSV and text files into it. As each file gets loaded, it can be appended to an existing dataset or forms a new one by default. Once these CSVs are in EDH, any of our duplicate detection algorithms (including chain duplicates) can be run on them.