The Software House team implemented an AWS-based data lake to automate xpate's accounting processes and secured $18,000 in funding through AWS' Proof of Concept program.
xpate's financial analysts were spending over an hour daily on manual work - collecting, processing, and analyzing transaction data to catch incorrect records. The fintech needed to automate this workflow while taking advantage of AWS funding opportunities. After a successful payment gateway project, xpate's CTO wanted to extend the partnership to tackle this new challenge.
Partnership goal:
→ Build a data lake that automates transaction reconciliation and qualifies for AWS' Proof of Concept funding program.
xpate
xpate provides payment services across Europe and Asia through multiple products. Their latest solution offers an all-in-one payment and banking platform for e-commerce merchants in the UK, with plans to expand across mainland Europe. The platform supports everything from traditional credit cards to Apple Pay and Google Pay, with pre-built plugins and API integrations for merchants of any size.
INDUSTRY
Fintech
COUNTRY
United Kingdom
SERVICE PROVIDED
Data Lake Implementation, Cloud Engineering
Challenge
xpate's accounting team faced a daily grind. Financial analysts manually collected transaction data from multiple sources, processed it, and hunted for discrepancies. This took over an hour every single day - time that added up quickly when you're processing thousands of transactions.
The company needed a way to pull data automatically from several different sources. But this wasn't just about saving time. xpate wanted to monitor success and failure rates for financial transactions — a metric that was critical for improving their reporting accuracy.
CTO Ruben Vosmeer saw an opportunity. As a long-time AWS customer, xpate could apply for AWS' Proof of Concept program, which rewards businesses for embedding Amazon services into their products.
Key requirements:
-> Automate daily manual data collection and analysis
-> Integrate data from payment gateway, ledger service, and authorization service
-> Enable real-time monitoring of transaction success/failure rates
-> Qualify for AWS Proof of Concept funding
-> Build on AWS infrastructure to maintain cloud commitment
:quality(75))
xpate’s financial analysts spend over an hour on manual work daily to collect, process, and analyze transaction data in search of incorrect transaction records
Solution
The Software House built an AWS data lake that automatically pulls transaction data from all three sources and feeds it into a single dashboard.
The solution centered on AWS services that could scale with xpate's growth. Lambda functions handle the data ingestion, AWS Glue processes and transforms the data, and QuickSight delivers the insights through customizable dashboards. This setup eliminated the manual work while giving xpate's team real-time visibility into their transaction metrics.
Process
Team:
The Software House engineering team (same team from 2022 payment gateway project)
xpate's Engineering Consultant
Technology Stack:
AWS Cloud
Lambda + Node.js
Serverless framework
AWS Glue + Scala
AWS QuickSight
Aurora Serverless
Amazon S3
How the data lake works
The architecture follows a clear pipeline:
Data ingestion: Lambdas written in Node.js pull data from external APIs and push it to S3 data lake ingestion buckets
Data processing: Changes to S3 trigger AWS Glue, which collects and maps data to an Aurora Serverless database that serves as the data warehouse
Data storage: AWS Glue saves the processed data back to S3 in parquet format for long-term storage
Data visualization: xpate's team reviews insights through AWS QuickSight dashboards
Learning Scala on the job
AWS Glue runs on Scala - a language The Software House team hadn't used before. But it was the right tool for the job, so the engineers learned it. Once they had Glue running, they hit a roadblock: Aurora Serverless didn't have native integration with AWS Glue. The team had to write a custom query that allowed Glue to push data into Aurora. Problem solved.
Securing the AWS grant
The Software House applied to AWS' Proof of Concept program on xpate's behalf for two parts of the project. The program rewards businesses that commit to using AWS services long-term. Once the data lake was operational, AWS paid out $18,000 to xpate — covering a significant portion of the development costs.
"The team is highly qualified and meets my expectations.
Additionally, I appreciate having a certain level of control during the process. The Software House is great at merging their team with ours and co-delivering.
Recently, The Software House has made their teammates more resilient to uncertainty."
Ruben Vosmeer
CTO at xpate
Read also:
→ xpate cut payout time for e-commerce merchants from 7 days to minutes
Outcome
xpate now sources first and third-party data into one dashboard automatically, saving dozens of hours in financial reviews every week.
→ $18,000 secured in AWS Proof of Concept funding
→ 1+ hour of daily manual data reconciliation work saved
→ Automated transaction monitoring across payment gateway, ledger, and authorization services
