DESIGNING UNIFIED PLATFORM TO BRING TOGETHER CUSTOMER JOURNEY DATA FOR IMPROVED DECISION MAKING

Know More

BUSINESS CHALLENGE

We analyzed the issues with the current data solutions

Airlines' daily operations are fueled with data from various functions (Reservation, Ticketing, Departure Control and Loyalty), leading to multiple data sources (GDS, OTA, Travel suppliers, Codeshare/Interline Partners, ATO/CTO office, Airport Ticket/Check-in counters and Loyalty Management systems). Some of the common challenges faced by airlines due to this include –

The internal repositories often operate in siloed fashion, resulting in lack of unified view of customer journey
Due to the rigid schemas and tight coupling with the core passenger systems, the Operational Data Store (ODS) significantly increases the implementation timeline, with every new data domain added to the platform
The analytics engine is not in sync with the ODS, thus defeating the purpose of real-time data analysis
The cluster contention and lack of scalability in the applications degrades the performance of technology solution
The cost to build and maintain an integrated data platform is exponential, and often does not justify the ROI

The drawbacks of existing airlines data solutions have come to the fore during the current pandemic, drastically resulting in huge ramp-down in airlines business operations.

THE SOLUTION

We offered a unified, flexible, scalable and cost-effective Mphasis Airline Data Platform

Mphasis brings together knowledge and deep expertise in the airlines industry, to create a data platform that serves both operational and analytical needs of the airlines. The platform uses open standards, purpose-built databases & serverless architecture, following well-architected best practices, reference architecture and native services of AWS. Our solution helps in -

Building basic data-as-a-service with the key tenet of separating storage from compute for most important domains such as Passenger, Flights and Loyalty
Building the Passenger, Flight and Loyalty data ingesting from Passenger Services System (PSS) in real-time through AWS Kinesis Data Streams, and capture, transform and load data streams via Lambda transformations into data lake (S3) and ODS (DynamoDB)
Creating a simplified Change Data Capture (CDC) pipeline for record-level insert, upsert and delete using Apache Hudi available within Amazon EMR service. Data sets managed by Hudi are stored in S3 using open storage formats, while integrations with AWS Glue Data Catalog give near real-time access to update data using familiar tools. Hudi manages the underlying data stored on S3 using Apache Parquet
Delivering microservices using AWS Lambda functions and events for DynamoDB ODS
Data analysis through AWS Athena using the data lake, and providing better visualization through dashboard using Amazon QuickSight
Building standard AI/ML models based on client use cases
Generating test data for customer journey in the OTA format
Adhering to IATA One Order services

THE OUTCOME

Three measurable results and multiple benefits achieved

Average infrastructure cost savings up to 60% annually

Average IT resource cost savings up to 50% annually

Business outcomes delivered within 2-3 weeks after a new information request comes in

Faster deployment within 6 to 8 weeks of integration lead time

System uptime increased from 99.90% to 99.97%, limiting the downtime to only 50 minutes in a year

There are several other benefits including faster time-to-market, reusability of the code, among others that can be seen with Mphasis Airline Data Platform.