Getting started with Amazon SageMaker JumpStart

An airline company with a loyalty program often encounters customers who signed up for the program multiple times using different email addresses but similar first names, last names, addresses, or phone numbers. If the airline company uses ML matching in AWS Entity Resolution to deduplicate 100,000 loyalty program member records by name, address, and phone number, and if these 100,000 member records are each stored as 3 rows of data in the airline’s Amazon Simple Storage Service (Amazon S3) data lake, then processing those 3 rows of data for 100,000 members will count as processing a total of 300,000 records in AWS Entity Resolution for a total cost of $75.

Airline: 300,000 records processed per month using AWS Entity Resolution

Note: Automated incremental processing is not available when using ML-based entity resolution.

A retail company can use AWS Entity Resolution to input dispatch information from different product suppliers to find common stock keeping units (SKUs) between them and improve understanding of supply chain efficiency. Each dispatch record can contain input data such as product SKU, product name, location, facility name, and manufacturer code. If the retail company has 15 suppliers with each of them sending 40,000 dispatch records, then the retailer can process all 600,000 records through AWS Entity Resolution rule-based matching to match and link unique product entities. In this example, the retail company will be charged $150 for processing 600,000 records. 

Retailer: 600,000 records processed using AWS Entity Resolution

If you use rule-based matching scenarios frequently, you can initiate automatic incremental processing so that as soon as new data is available in your S3 bucket, AWS Entity Resolution reads those new records and compares them against existing records. This keeps your matches up to date with any changes in S3 data. When using automatic processing, you will only be charged for incremental processing. 

A retail company can use AWS Entity Resolution to input dispatch information from different product suppliers to find common SKUs between them and improve understanding of supply chain efficiency. Each dispatch record can contain input data such as product SKU, product name, location, facility name, and manufacturer code. If the retail company has 15 suppliers with each of them sending 40,000 dispatch records per month, then the retailer can process the new or incremental 600,000 records per month through AWS Entity Resolution rule-based matching to match and link unique product entities. After the initial workload, the retail company will only be charged $150 for the processing of 600,000 incremental records per month and compare them against the existing 600,000 records processed the month before. The retail company is not charged again for processed records from the previous month.

Retailer: 600,000 incremental records processed using AWS Entity Resolution

If you use rule-based matching scenarios, you can initiate near real-time resolution. AWS Entity Resolution reads those new records and compares them against existing records in near real-time. This enables you to quickly match records for real-time use cases such as tailoring product recommendations, personalizing guest experiences, improving patient care, or identifying fraudulent transactions. When using near real-time matching, you will only be charged for incremental processing. 

An airline can use AWS Entity Resolution to match records between online booking transactions and their frequent flyer database in near real-time to find loyal travelers between them and tailor recommended experiences based on their preferences. The airline processes 50 booking transactions per second and each record can contain input data such as name, flight number, origin, and destination. If the airline has 100,000 booking transactions per month, then the airline can process the new or incremental 100,000 records in near real-time (processing the 50 records per second) through AWS Entity Resolution rule-based matching to match and link frequent flyers. After the initial workload, the airline will only be charged $25 for the processing of 100,000 incremental records and compare them against the existing 900,000 records previously processed. The airline is not charged again for processed records from the previous job.

*You can enable near real-time matching through the AWS Entity Resolution Generate Match ID API, AWS Management Console, or CLI. Learn more.

Retailer: 100,000 incremental records processed in near real-time using AWS Entity Resolution

An e-commerce company with 1 million customer records can use data service provider matching in AWS Entity Resolution to link and translate these records with common industry IDs and provider data sets, so they can more effectively reach their customers across marketing channels such as search engines, social media networks, and digital advertising. These 1 million customers already interact with the company’s various channels including web, chat, email, and applications. Using this matching workflow, the company can connect and translate their records with industry IDs including RampID and Unified ID 2.0 to better understand and service their customers. Companies that use this matching workflow will be charged a data service provider subscription cost through Amazon Data Exchange (ADX) in addition to the record processing cost through AWS Entity Resolution.

e-Commerce company: 1,000,000 records processed using AWS Entity Resolution

A travel company wants to personalize their marketing and advertising campaigns to with highly-relevant messages to attract and delight their desired audience. The travel company does not have a holistic understanding of this audience so they license 500,000 records from data providers to enhance their own records, such as demographic attributes that help them better understand the audience needs. Using the data service provider matching workflow in AWS Entity Resolution, the travel company can translate and enhance their records with industry IDs such as TransUnion TruAudience, and append additional columns to their records including demographic attributes.

Travel company: 500,000 records processed using AWS Entity Resolution