Most of the world’s largest banks want to innovate, using the terabytes of data they hold to improve customer service and deliver new products.
They need to: Pressure from fintechs and startups that have minimal technology debt as well as their peers is mounting, regulators want more insight from market datasets, and shareholders want to see return on the substantial technology investments that banks have made in recent years.
Yet delivering data-powered innovation is not easy in large organisations – and certainly not when it is locked into deeply entrenched legacy applications. One example is Sybase Adaptive Server Enterprise (ASE).
Sybase’s end of support triggers a rethink
A powerful relational database management system, now owned by SAP, Sybase took Wall Street by storm through the 1990s and remains deeply embedded in many organisations. But mainstream support and patching for the software ends in 2025, and most users are exploring alternatives.
The ubiquity of Sybase applications renders this a huge challenge. As the first client-server database that could handle large-scale financial services data workloads via a SQL interface, many mission-critical securities trading and other applications still run on the ageing software.
Keeping applications running on Sybase however essentially means that the data flowing through them is partitioned off – and interoperability with newer software is a huge challenge. One major multinational bank wants to see change. Working with Accenture it is embarking on an ambitious shift off its embedded post-trade applications, from ASE to MongoDB.
A knotted nest of stored procedures
Speaking to The Stack, Accenture Associate Director Guruprasad (Guru) Rao, said: “What we have noticed is that most trade and risk applications are on Sybase… it was easier for adding stored procedures seamlessly.”
(Stored procedures are, in short, precompiled subroutines used to add new functions and consolidate application logic, he explained.)
This means that such applications have sprawled in complexity over the years as banks add customisations; not all of them well documented, making migrations to alternative software even harder to achieve.
As Accenture’s Guru – who is leading work on a proof-of-concept to demonstrate how the migration to MongoDB Atlas could add value – put it: “For each [new] function, developers started adding one stored procedure. ‘Here’s a function I want to achieve with this application: add a stored procedure – and business logic inside the stored procedure.’”
Shiv Pullepu, MongoDB Industry Principal, Financial Services, added:
“Over the years it's common for any business-critical [Sybase] application to require thousands of stored procedures underneath it [and] become a humongous monster. It's stable – but it's not scalable for the future.”
Pullepu is a banking veteran himself; prior to MongoDB, he worked for the Royal Bank of Scotland and UBS Investment Bank.
Sitting alongside Accenture’s Guru Rao and speaking to The Stack, he said: “All the banks I worked in had Sybase. All my friends who work in finance still actively use and manage Sybase. It’s a good platform.”
“But it’s not good for interoperability and it’s not designed to work with AI/ML platforms. It’s not cloud-native, or good at providing support for analytics or search natively; [so] you end up creating or buying them.”
With Sybase’s similarities to SQL, it would be easier to migrate to a relational database. But the scalability, flexibility and battle-tested nature of MongoDB Atlas made it compelling to the bank’s technology leaders.
As Pullepu put it: “Atlas is not just a database or a persistence layer.”
“It’s a platform. It comes with all these embedded features like search, analytics, flexible data model in a single unified platform.
It will save a lot of money [and open up new capabilities]” he told The Stack, highlighting Atlas's powerful “queryable encryption” capabilities.
Sybase to MongoDB: A POC takes shape
This particular bank had explored modernising Sybase-based post-trade applications itself, but had run into issues with its stored procedures.
Accenture is now untangling a knot of “stored procedures” to deliver a proof-of-concept migration to Atlas – which with the bank’s technology leaders it chose for its speed, scalability and multi-cloud capabilities.
Accenture has a history of success here with migrating legacy applications to MongoDB’s noSQL document model. It had previously migrated “around 200 to 225 legacy applications which were on mainframes” to MongoDB Atlas for an entertainment sector client, Rao told The Stack.
“It was a big success for our data practice and gave us the level of confidence to say ‘let's explore what we can do here.’” he said.
To get the proof-of-concept started, Shiv Pullepu added that MongoDB presented the bank with numerous customer stories, highlighting a variety of mission-critical financial services apps that utilise MongoDB Atlas, including both pre-trading and post-trading applications across the globe.
A careful shift
Work on the proof of concept is proceeding carefully, they said.
Asked about the unique challenges a Sybase migration entails, Accenture’s Rao said that shifting from a legacy SQL to a NoSQL database comes with a set of standard challenges: “The data models are completely different… the schemas are different… the integration patterns are different. When you go to NoSQL, you need to unlearn what you learned on a relational database, and try to query it in a completely different way,” he added.
There’s also the challenge of complex custom code and erratic documentation with many older applications, he added. “[For this particular post-trade application] when we looked at the stored procedures, we were blown away [wondering] how can somebody write these kinds of complex stored procedures for a simple function?” he said.
Join peers following The Stack on LinkedIn
Tackling this is a bite-at-a-time job: “We are reviewing one store prod at a time; understanding the business logic and rewriting it in a way which is easily comprehensible and easily migrated to the application layer.”
It’s undeniably a challenge. But Rao and Pullepu say that once the Sybase to MongoDB migration is completed, it will free the bank up to do a wealth of new things with the application data including with AI, reduce cost, and help it migrate to the cloud – a priority for the bank’s digital leaders.
The team are themselves exploring the use of AI in the POC too.
For example, they are using MongoDB’s Relational Migrator to understand the application’s logic, and write alternative code, as Rao puts it, “so that the data layer becomes nimble and lighter, and application layer owns the logic now – making it easy to migrate the data, and [the] application [itself becomes] nimble and faster and higher performance.”
It is, added Accenture’s Rao, “high stakes.”
He said: “If we can prove this it opens the doors for thousands of legacy applications that are sitting [on Sybase and other software].”
For the bank, it will enable critical insights from data flowing through some of its most important applications and open up the possibility to build completely new products and services on top of that data.
Delivered in partnership with MongoDB