Moneygator was a rapidly growing personal finance portal, finding the the best mortgage, credit card and personal loan deals for consumers.
Business was taking off with the company experiencing a rapid increase in traffic and signing affinity deals with other portals.
Problem was they were having growing pains... of the scalability kind. With a small number of users, the site would perform acceptably, but once more than a handful of customers would be online, the user experience became erratic with long delays in page loading being common as the number of users rose.
DataSurge was tasked with identifying the scalability problems, recommending a course of action and leading the in-house development team in implemention. Moneygator's business had to go on, despite their scalability problems, so an additional task was to increase scalability in the short term until the re-architected system was deployed. This we did, loosening the bottleneck from a handful of simultaneous users to several tens.
We identified fundamental architectural problems and set about re-architecting the site for high performance. Our first phase was to investigate four types of data access strategy recommended by Microsoft and set up a proof-of-concept for each to test for performance. Our "proof-of-concept" strategy was fully justified when we found a serious weakness in one of the strategies, but then discovered a far more successful way to deploy it.
Having defined the architecture, we set about coding, working with the in-house team to create a highly scalable and fast performing personal finance portal. This was based purely upon the latest Microsoft technologies at the time including SQL Server 2000, ASP, IIS 5, Windows 2000, COM+ and SQLXML.