EntityStream.com MARS project is designed to provide flexible matching microservice to enable you to compare, match, index and standardise records representing any type of object such as Person, Legal Entity, Contact, Employee etc.
Mars is a small but perfectly formed “microservice”, that allows you to present it with two records and have it index, standardise and compare them together so you can understand how similar they are in real life scenarios.
Mars doesn’t use traditional relational or deterministic technology, in fact it uses a rules extended probabilistic engine called “Identiza” that was developed by Robert Haynes over the past decade. This type of engine is often called “Fuzzy” Matching and has been the basis for many products like Informatica (tm) MDM and IBM Initiate (tm) technology, however the problems with these technologies are that they are incredibly hard to understand and also very complex to extend.
EntityStream Mars is a much simple way to process data, instead of relying on converting all data sources into one form so they can be compared and contrasted, Mars instead allows you to store the data in a closer to nature form, that doesn’t require any significant processing before it can be matched. Match rules and comparisons are done between heterogenous data structures and this means the chances of error is greatly reduced due to the lack of loss of data quality when you transform data.
MARS Configuration is simple - deploy from one of the four cloud or on-premise options:
Pull the latest docker image from https://hub.docker.com/r/entitystream/mars/ into a new Docker container Instructions
Create a Google Compute Engine from gcr.io/psyched-battery-225414/mars:latest using https://console.cloud.google.com/compute/ Instructions
Create a Azure container instance at https://portal.azure.com/#create/microsoft.containerinstances Instructions
Coming soon - Amazon EC2!