Accelerating the creation of AI-infused solutions in a hybrid environment

Share this post:

As a global leader in software for banks and financial services organisations, Finastra aims to bring generative AI (gen AI)-enriched solutions to its clients without limiting their options around choice of platforms.

Steve Moe, Head of Technology for the Lending business at Finastra, explains how a collaborative initiative between IBM, Microsoft and Finastra, using the IBM Build Partner program, is helping drive innovation and choice for AI.

Designing a structured approach to AI

Finastra always wants to offer innovative, future-proof technology solutions to our clients, and we were already five years into our own AI research when OpenAI kick-started the gen AI revolution with ChatGPT.

Suddenly, gen AI was a real tool, ready for mass use. In response, we rolled out a gen AI adoption program in three phases: educate, enable, execute. After a year’s worth of upskilling and enablement across our ~8,000-strong workforce, we reached ‘execute’ mode in March 2024. We now have a program we call “50 Flowers Blooming” — that is, 50 use cases across internal and customer-facing solutions, that Finastra plans to bring into production in the coming nine months.

We’ve taken our time to build rigorous, enterprise-ready workflows around AI because our industry can’t afford to work with anything less. Our gen AI product engagement roadmap covers 16 steps, including securing internal legal and compliance sign-offs through a PoC stage, validation testing, governance, and finally release to the global Finastra employee base or, indeed, to our clients.

Keeping an open outlook with a hybrid approach

A cornerstone of our gen AI approach is that we won’t standardise on a single technology. We want to keep our options open so that our clients can choose the technologies that fit best with their own infrastructure and preferences.

It’s similar to our approach to cloud, where the ethos has always been: “Build it once, deploy it to multiple clouds.” Our first non-functional requirement in gen AI is: you can’t pass into production if you’re just using a single model from a single provider.

The reason we’ve chosen IBM alongside Microsoft in our collaborative approach to gen AI is that we need providers with proven, enterprise-class products. The heritage of IBM watsonx is certainly impressive, but what appealed to me most as a CTO was the governance and validation that IBM provides in its AI models. I also believed that IBM AI technologies would offer real value to our back-office technical people, whereas other AI technologies seem to be aimed more at the consumer market.

The IBM Build Partner program is helping us to create a package of 7 MVPs [minimum viable products] with gen AI features. In addition to the watsonx technology, we’ve gained access to IBM services and a global ecosystem of experts for a fast start.

Rapidly bringing new AI-enriched tools to market

We’re using multiple offerings within the watsonx portfolio — including watsonx.ai, watsonx.governance, watsonx.data, watsonx.discovery and watsonx Assistant — to build our MVPs.

One of those solutions is a persona-based, multi-modal learning tool infused with AI that provides more personalised experiences for people training to use our Loan IQ lending platform. There’s a big drive for certification in our market, and a shortage of people with the right skills. Using AI to help accelerate training could deliver significant benefits.

Taking a serious approach to safety and security

Working with IBM has been positive. We got IBM’s full attention when we came to them with a business problem, which enabled us to get our MVPs completed within a matter of a few months.

We also appreciate IBM’s thinking around security, governance, and ethics in AI. The watsonx portfolio reassures us that IBM has meticulously considered the safety and security aspects of gen AI from an enterprise perspective. This results in a toolset we are comfortable using within our own rigorous quality-control process.

Looking ahead, we’re excited to leverage the watsonx portfolio to create solutions that deliver tangible value from AI for both internal teams and clients.

About Finastra

Finastra is a global provider of financial services software applications across Lending, Payments, Treasury and Capital Markets, and Universal (retail and digital) Banking. Committed to unlocking the potential of people, businesses and communities everywhere, its vision is to accelerate the future of Open Finance through technology and collaboration, and its pioneering approach is why it is trusted by ~8,100 financial institutions, including 45 of the world’s top 50 banks.

About the author

Steve Moe is a senior technology leader in the financial services industry. He has worked in global leadership positions across diverse disciplines within banking for more than 25 years. Drawing on his extensive experience on both the FI and IT partner sides of the table, Steve is currently Head of Technology for the Lending business within Finastra.