We have crafted solutions for industry leaders, combining years of expertise with agility and innovation.
An insurance business has gone global by creating a partner network across the world that sells its services. Its insurance partners plug into the businesses’ exchange system which calculates customer premiums, cash bids, points and rewards.
The better the system and service to its partners, the greater its global success.
However, its partner network has been growing so rapidly that the system has been slowing down. The business had to alleviate it growing pain, and quickly.
The business’s partners have been experiencing growth in clients which means more daily data that needs to plug into the system. Clients are constantly inputting information that affects their points. The faster the turnaround in data, the better the customer experience.
The batches being uploaded into the system have become larger and larger. The whole system comprises of several microservices which processes the batch and then passes the data onto the next part of the system. If one part of the system slows down, the whole system slows down.
With growth expecting to only increase, the insurance business consulted Synthesis Software Technologies, an innovative software development company, to optimise the system and create resiliency.
The Synthesis team was asked to solve slow processing in a one part of the system.
“The performance of their system is measured by how long it takes to process batch jobs of approximately 50 000 records of data on Spring Batch. I decided to do what many consider as a last resort – dive into the training manual and read it end to end,” explains, Kim Banga, Synthesis Software Engineer.
“With extensive research, “I managed to rebuild the processing system to take full advantage of Spring Batch’s ability to run multiple batch processes simultaneously and this led to substantial improvements in the system.”
Before the system was not being worked to capacity. Now the system runs at 100% capacity through multithreading – accelerating the processing of data.
This led to a 30% improvement, but the Synthesis team were still not satisfied with this. The insurance business needed more with their expected growth. The team experimented with different service resource configurations to maximise the results.
The team also had to ensure that partners that wanted to quickly process a small batch could do so as well. There had to be a balancing act that processed the batches as fast as possible without overextending the system or underutilising it. They implemented parallel batch processing and found the perfect number of batches to process at once through testing the system.
“I have been astounded by the results,” exclaims Banga. Through deep research and testing, there are now performance improvements where a part system runs 75% faster. This slowest part of the system is now the fastest part – making a massive difference in the system as a whole.
The insurance business is now shifting its focus to other parts of the system that once seemed fast but are now slow compared to the recent improvements – the business has raised the bar.
“For this particular part of the system, processing of 50 000 records dropped from an average of 6 hours 30 mins to an average of 1 hour 25 mins. A user now only has to wait an average of 4 hours for a 50k batch to process, down from 9 hours. We are now more than 50% faster even before optimising and stabilising the other parts of system.”
The time it takes for the system to ingest, process and return the information to the partner network has drastically declined and improved fold.
This improves the partner experience as well as the customer experience. Points, rewards and premiums are now available 50% faster on the partners’ apps and this is just the start.
The business has seen what is possible for its partners and their customers and is determined to keep raising the bar. The business is set for further expansion.