A-Team Group’s Innovation Briefing series got off to a great start in London last week with an event dedicated to innovation in cloud. The briefing included practitioner interviews, provided insight into how to modernise data infrastructure, and reviewed technologies driving change.
The case for data mesh
Kicking off the briefing, and raising several delegate questions, a practitioner interview discussed the merits and limitations of data mesh as a means of data infrastructure innovation. Duncan Cooper, head of OMNI Digital Services at BNY Mellon asked the questions, which were answered by Sunny Jaisinghani, technology platform owner – data mesh, formerly HSBC, and Simon Massey, data mesh lead, formerly HSBC.
The speakers described data mesh as a gamechanger with the ability to deliver on the dream of maximising all data, create data products in as little as eight weeks, and address scale and agility challenges. It is, they said, not about the tech stack, but about a way of working with data.
This new dimension in data architecture is about moving away from a technology view of data architecture to a more holistic view. Key to success is a 25% focus on engineering and architecture, and a 75% focus on design. Culture is the secret sauce, with small DevOps teams iterating quickly to deliver value faster than traditional data builds.
Aligning data mesh to use cases and opportunities to scale is key, noted the practitioners, pointing out that use cases that are building blocks in an organisation provide value that can be attributed to the mesh, avoiding the traditional challenges of winning buy-in and budget for every project..
It is not all plain sailing, however, with Jaisinghani adding a caveat that mesh is not a solution for all things data. It needs to be ringfenced, he said, and is often used alongside data fabric to match a firm’s technology estate.
Differentiating data fabric and data mesh, the speakers described data fabric as the result of abstracting datasets to gain quick value, but suggested this mechanism is already in use and is not sustainable. Data mesh pushes data onto a shared platform for wide consumption and the creation of data products.
Moving to cloud
Increasing volumes of data resulting from regulation, the need for near-real time data flows across the front, middle and back offices, and a desire for agility that can help financial institutions improve product creation and speed to market are some of the reasons firms are transitioning from legacy data infrastructure to cloud, according to a panel discussing the move to cloud at the innovation event.
The panel was moderated by Colin Gibson, senior advisor and regional advocate at the EDM Council, and included Hugh Davidson, head of data management at Man Group; Oli Bage, distinguished engineer, head of architecture, data and analytics at LSEG; Dan Seal, SVP, streaming analytics at KX; and Ihyeeddine Elfeki, trading and risk management solutions lead at Luxoft.
The speakers addressed the current state of migration to cloud in capital markets and ensuing changes, including fundamental changes to the technology stack, and data use and ownership. They also noted the need for a balance between current business and future state as cloud is implemented, and agreed that reengineering technology for the cloud is not trivial.
That said, the objectives of modernising are many. An audience poll showed 77% of respondents citing superb agility and analytics as their key objectives, with 62% citing significantly improved data quality, 46% reduced costs of data and data management, and 38% support for advanced analytics and machine learning programmes.
The speakers touched on cloud capabilities such as the ability to scale, achieve agility, and vastly improve analytics as key objectives of modernisation, but also on the need for fine grained, automated controls if firms are to make a successful transition to cloud. They also considered a hybrid approach of using the right tools for the right job to achieve parity across on premise and cloud systems.
The challenges of modernisation include reorganising technology and data at the same time as addressing business challenges, and ensuring the goals of moving to cloud – particularly elastic compute, scale, agility, productivity, and business objectives – are actually realised.
Data and AI innovation
A final panel session discussed the role of data and AI innovation in business transformation, again agreeing that change is a large challenge, but if managed well, can deliver sizeable benefits.
The panel was moderated by Sarah Belsham, data, analytics and insights partner at RSM, and joined by Peter Jackson, director, chief data and analytics officer at Carruthers & Jackson; Nav Hira, head of data management at Bond Radar; and Andrea Nardon, CIO at Creed & Bear.
The problem of top-down culture causing reluctance to change at many financial institutions was raised and addressed by Jackson, who described his experience of moving data from mainframe systems to cloud and resulting business changes. He said explaining the risk of doing this to senior management is difficult, but must be done, along with how the risk will be mitigated. Finding the right skills to build an operating model and make it work is also challenging.
Moving on to AI, the panel highlighted the potential of the technology, as well as ethics issues it presents and the need to think carefully about doing things ‘just because you can’ – one of many conversations that continued among speakers and delegates after the event closed.
Subscribe to our newsletter