Unified Profiles for Travelers and Guests on AWS

Connect traveler and guest data to unlock insights and drive personalization

Overview

Unified Profiles for Travelers and Guests on AWS helps you automatically source, merge, remove redundant data, and centralize traveler and guest information from across your enterprise. It offers a single data source for marketing, sales, operations, and customer experience leaders, providing the data they need to create a central view of the customer. A seamless and personalized customer experience helps deepen the connection between the customer and your brand, and with a few clicks, customer information is aggregated from multiple internal systems of record, producing a unified customer profile in both batch and near real-time. Now, you can achieve a faster and a deeper understanding of your travelers and guests to deliver a tailored, more personalized customer experience.

Benefits

Easily connect travel and hospitality industry systems

Integrates with your property management systems and customer data platforms. Also, a detailed process plan is provided for all common travel and hospitality data sources, such as bookings, loyalty profiles, hotel stays, and customer service interactions, minimizing the data transformation work to integrate this solution.

Resolve traveler identity and reduce duplicated data

Use a constellation of rule-based and artificial intelligence (AI)-based Identity Resolution capabilities. This solution's modular architecture also allows for custom-matching logic and can be integrated with AWS Partners.

Secure storage and permission management

Sensitive traveler data is stored in an AWS specialized customer data store and located in your AWS account. This helps you deploy a permissions system with strict access controls to traveler data.

Capture traveler changes in near real-time

Provides near real-time updates so that your customers can react to all changes within their traveler profiles, such as booking cancellations or hotel checkouts. Leverage a complete view of your travelers and guests for timely and personalized communications.

Technical details

You can automatically deploy this architecture using the implementation guide and the associated AWS CloudFormation template.

Unified Profiles for Travelers and Guests on AWS | Architecture Diagram

Step 1
A batch ingestion module allows you to send data to Amazon Simple Storage Service (Amazon S3), run extract, transform, and load (ETL) jobs with AWS Glue, and ingest data in highly parallelized AWS Lambda functions into the solution's core storage module (interaction store).

Step 2
A real-time ingestion module allows you to send real-time data to Amazon Kinesis Data Streams. Data is then transformed using the same transformation function used by the AWS Glue jobs in the batch ingestion module and ingested into the solution's interaction store.

Step 3
The core storage module (interaction store) is an Amazon Aurora Serverless cluster that ingests transformed records (interactions) in a custom-built transactional data store while maintaining consistency and data lineage. This module also optionally caches every customer profile into an Amazon DynamoDB table for low latency access.

Step 4
The interaction stitching module indexes every interaction into a dedicated Amazon Aurora table and runs lightweight match queries against the index table to identify interactions that match a set of rules (configured in the solution frontend).

An AWS Fargate cluster allows you to start a large number of parallel tasks that fetch interactions from Aurora. You can then replay them against the index whenever rules change. This module sends interactions identified as a match of the same profile to a merge queue for consolidation.

Step 5
The Amazon Connect and AI-identity resolution module receives relevant consolidated profiles from core storage and caches them in an Amazon Connect Customer Profiles low latency store.

Identity resolution is run by Amazon Connect Customer Profiles, which identifies matches and sends them to an S3 bucket to be processed by the solution (and merged above a threshold). Once in Amazon Connect Customer Profiles, you can leverage the Amazon Connect integration and proactive engagement features.

Step 6
The change processor module receives all changes made to a profile and sends them downstream to an S3 bucket. Amazon EventBridge allows you to easily integrate downstream applications.

Step 7
The administration portal module provides a user interface and a REST API, allowing you to configure the solution and search and retrieve profiles. This web portal can be integrated into an Amazon Connect third-party application framework to be used by customer service agents.

Step 8
An error management module is composed of 13 Amazon Simple Queue Service (Amazon SQS) queues strategically placed throughout the solution along the upstream and downstream data path to gather all errors (transformation, ingestion, downstream event creations, and identity resolution).

These errors are processed and indexed into a DynamoDB table and are available in a sorted real-time log through the admin portal, making the solution easy to monitor. The solution also includes a custom Amazon CloudWatch dashboard and a set of custom metrics using a CloudWatch-embedded metric format.