SNIA Experts on Data
Listen to interviews with SNIA experts on data who cover a wide range of topics on both established and emerging technologies. SNIA is an industry organization that develops global standards and delivers vendor-neutral education on technologies related to data.
SNIA Experts on Data
The Future of Financial Services: AI, Big Data and Data Architecture
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Ready to unravel the complex world of data architecture in financial services? Join us for this conversation with SNIA member Parviz Peiravi, Global CTO for Financial Services at Intel, as he dives deep into the dynamics of data structure and its impact on the financial services industry. The interview takes a close look at the challenges faced by today's financial organizations, including the limitations of traditional data architecture, data silos, cultural barriers, and security issues that impede the adoption of AI and big data technologies. Parviz also sheds light on the implications of data regulations, underscoring the necessity for a robust, modern data architecture to ensure compliance, and offers the steps that financial institutions can take to foster this culture, touching on best practices for data quality, data ops, and data observability.
SNIA is an industry organization that develops global standards and delivers vendor-neutral education on technologies related to data. In these interviews, SNIA experts on data cover a wide range of topics on both established and emerging technologies.
About SNIA:
Welcome to the SNEA on Data Podcast. Each episode highlights key technologies related to handling and optimizing data.
Speaker 2I am your host, mark Brown, representing SNEA, and I am sitting with Parvus Parvati, who is and you can correct me if I pronounce your name from there Parvus who is Global CTO for Financial Services at Intel and a member of the SNEA community. I have been looking forward to this conversation with Parvus to talk about the challenges of data architecture when it comes to global financial trading and the financial services industry. So, no more ado, parvus, would you like to introduce yourself? You can correct me, honey, incorrect names.
Speaker 3No, you're perfectly fine. So my name is Parvus Parvati. As mentioned, I am Global CTO for Financial Services Solutions at Intel. Basically, our team works with largest financial institutions worldwide in working with them in collaboration, designing different type of solution for AI, for data management, etc. Super.
Speaker 2So many of our members would probably be interested in the topic and I would love for you to expand and explain why traditional data architecture is not really up to par or sufficient for the evolving needs of the financial services industry. Can you tell me what the challenges are, what people are facing in the call phase?
Speaker 3Yes, so we have been in an evolution of data architecture over the last 40, 50 years, but various load, unfortunately. The old traditional data architecture was very close to organizational architecture that was developed by different institutions, enterprises etc. And that basically created a rigid structure, an inflexible structure that could not easily use data from different groups and business units and organizations even within one basic institute, and that inflexibility really creates significant problems. However, during the years, we tried to address that with basically using Band-Aid and that solved some of the challenges temporarily. For us that create more issues.
Speaker 3I call it an organized chaos basically, and that type of architecture really didn't look at data from data value data as a product.
Speaker 3It looked at data as a byproduct of different activities and therefore make it much more difficult for using data in different organizations for different use cases, etc. Because it was strongly attached to a specific use case and creation of the data was created. So this old architecture really it's a major barrier for adoption of newer technology. Nowadays and this is more obvious for everybody, especially with we are saying that big data with artificial intelligence we are seeing generated AI recently is a spread in everywhere, and all of this technology relies on high quality data, volume of data and accessibility to data in real time. So that really is forcing and driving adoption of new ways to deal with old data architecture and creating the modern data architecture that is going to help them to actually use data and value data as it should be. And in last stage of that, hopefully, we are looking at data as a product so that way we can utilize it even more effectively, because we treat it just like a product, with risk associated with product, security associated with product and so forth.
Speaker 2So what are today in everyday financial terms? What are the key challenges that financial institutions face today when it comes to managing and utilizing data effectively?
Speaker 3So one thing we have seen in the last six, seven years at least we are seeing that multi-national regulations. So we have EU regulation, we've got Chinese regulation, india has regulation, us has its own. Every country has its own regulations and majority of, especially financial institutions that they are multi-national. They are dealing with all of those regulations and so regulations that are directly affecting the financial institutions today. One of them is data sovereignty. That means you can move data from specific countries outside of that country. The other is data privacy reaches below. We are seeing a specific implementation of that with EU with GDPR, and we have seen it with California recently CCPA, and so we're going to see a lot more of that coming up. So that's number one challenge that we are facing. Secondly is really we are talking about data silos and that's come from the old architecture attached to a specific organizational structures and inflexibility of make the data accessible and being in institutions being able to actually take full potential of the data, which is required.
Speaker 3New data architecture and looking at data and, as I mentioned, like a product, developing modern data governments and associated policies and infrastructures. We're talking about massive data, reference, data management, metadata management, incorporating data observability, etc. As part of infrastructure that manage the environment. Cultural barrier is another aspect of it. Especially, we're talking about data literacy from executive level all the way down to employee level Understanding data what is data, value of data and importance of data in daily life of an enterprise and this is really become an important factor as we go forward.
Speaker 3Security architecture of course, it's something that we are seeing. Security has always been on top of every CTO, cisos and CEOs as well. With the recent activities that we are seeing that the require new type of architecture. You cannot really live with the security architecture of the past and trying to move forward and providing services that the customers are looking for or internal business unit are also looking for. So this is important to adopting security architectures. That is multi-tiering. It's basically looking at the data, data security, the cyber security and also looking at it from adopting a concept of zero trust. And I'm not saying we mentioned creating the metric for data evaluation. So data value, data quality those metrics is going to help organization to deal with the challenges of adopting new architecture, et cetera.
Speaker 2So you mentioned GDPR and CCPA, which are regulations, and I'm wondering it's kind of a two-fold question. So there's security concerns, certainly, and I'd be interested in what the top security concerns are for the day-to-day operations of financial transactions. And also, from how do data regulations such as GDPR and CCPA impact the way financial institutions are doing it today? Is that a big deal, as well as the security concerns, and how is it impacting them? Do you think?
Speaker 3Oh, absolutely so. It's not just a regulation, but that infrastructure needs to be in place to be able to comply to those regulations.
Speaker 3So it's not just looking at it from a hypothetical approach. It's a practical approach that we need to look at is that do I have infrastructure in place to provide the accessibility to data to a customer? So, if we are talking about, basically, gdpr, it's enabling customers to make a decision about their data. Now, is financial institutions infrastructure providing such capabilities to customers? And that goes back to if they do how they do it, do they provide security on top of that, because you are enabling customers to make a decision on their privacy perspective? But is this mechanism that you have in place is secure enough that provide that level of access to data? And we are seeing other things that in Europe, like open banking, data sharing they all require infrastructure to comply with that. So, yes, it's a major challenge for financial institutions to comply with those regulations if they don't have right infrastructure in place.
Speaker 2Okay, so in terms of that, then, you mentioned as well silos, so I'm just wondering how are institutions handling that silo data when they're moving to say it was a lot of movement to cloud based solutions, and how is that transforming the way financial institutions handle and process data?
Speaker 3That is why Majority of financial institutions start looking at the modern data architectures and, associated with that, modern data governments. Because when you are moving data from on prem to a cloud, then that become a complex infrastructure. Number one the governance, because it become more complicated. Access control become more complicated. In the past, on prem, every institution managed the entire aspect of the security, access control, etc. Now you are sharing that because big part of that is going to be handled by crowd infrastructure management team and cloud solution provider combination. So take into consideration that moving data to cloud is not just moving part. You have to build that policy, access control and governance, in addition to collaborating with cloud vendors, both the solution vendors and infrastructures that ensure that meeting your security requirement in all aspect of modern security implementations.
Speaker 2Okay, so just going back to what you said, you mentioned AI and machine learning. I'm just wondering what you see is emerging with technologies like AI when it comes to decision-making with financial services, and what, for instance, for the membership who are watching this podcast for us and the member who may not even be in the financial services sector but maybe involved in AI, what should they be thinking about when it comes to this?
Data Management Trends for Financial Services
Speaker 3Well, frankly, what we are talking about here today applies to every enterprise. This is not just financial services, but first my focus in financial services will be specifically target that. But in general, ai is going to be everywhere and it should be an initiative for every enterprise that looking at AI everywhere concept and understanding how that can actually happen. Because we talk about number one aspect of it data. If there is no data, forget about AI. There is no intelligence, because you need the good quality data to be able to actually utilize it, analyze it, which is the decision-making with artificial intelligence, big data, machine learning, different kind of practices of analyzing data.
Speaker 3So you have to build a modern data architecture to be able to feed data to different type of AI that your organization will meet and, trust me, automation and intelligence it's going to be needed in almost every aspect of your enterprise for us, from business process implementations, which we talk about, robotic data processing, and we are talking about AI in decision-making for executive, ai in automated decision-making for many a business unit that they're going to use it, analyzing data to providing effective, personalized services to your customers. So, internal, external automation, cost reductions these are all part of the use cases that I've seen in financial services and outside of financial services that they're going to use AI. So be ready for building a solid data architecture can feed those AI type of technologies that are going to help you either grow your business or effectively competing in the market and reducing the cost of operation. So these are the top of the line for CEOs that I've been I've spoken to that they would like to take advantage of AI.
Speaker 2Right. So that's interesting because I could see then there's cultural business challenges there. So what steps should financial organization take to foster a data-driven culture among the financial services, as well as working with other organizations?
Speaker 3So the steps that need to be taken is number one, a step back. Take a look at your data architecture as is today. It is a very difficult and challenging task to completely redesign everything at once, so it is important to understand the capability of your data architecture and issues and problem with that and then start, one by one, addressing those in pieces. For example, some institutions today and I've seen this and this is a public data with JP Morgan Chase they start adopting data mesh approach. Data mesh is an architectural approach, not the technology that look at the data as a product and build the environment that support that architecture and offer it. It's a fully distributed data architectures, distributed governance of data as well, and that is helping institutions to address some of the challenges that we are seeing Now.
Speaker 3This is not end of all. This is one of the solution to the problem we are seeing at the moment, but the important part is understand the current environment, understand what would be the changes you need to do to get the best out of your environment as is, and then start building up a roadmap to the implementation of modern data architecture that require both new governance as well as new data architecture, implementation in combination and the technology associated with that. So this is what we talked about people process technology before, but this is people process technology and data now. So this four factor is going to help you to start understanding what framework, or frameworks, actually is going to help you to address the challenges of your organization. There is no one solution for everybody, but there are multiple frameworks and approaches available that could help them to build the new modern data architecture, modern data governance environment for their enterprises.
Speaker 2Okay, so for the Senean members, what would you tell them would be some of the best practices to start off with to ensure data quality for their financial services or even to get them in a position where they can work and work with other organisations around that where financial transactions and financial services are required.
Speaker 3So having a solid data lifecycle management is a key aspect of basically dealing with that Data in association. There are six or seven, so it's not a short answer, unfortunately, but data lifecycle management and data risk management, backup recovery with the new again, you have to look at backup recovery with the new requirements in enterprise. We're talking about just batch oriented real time, near real time, the streaming. These are all key factor of understanding new technology and new approach in backup recovery. Data access control is especially very important. Data storage management since we're talking about the Senean members, the new storage technology to actually be in a foundational infrastructure for streaming data management for, again, near real time data management, etc. Data bridge prevention, confidentiality, integrity and availability of all those infrastructure and data protection policies that goes with that data governance that we mentioned. These are the key factor of new data management in general within an enterprise and I think Seneal contributed a lot into many of those aspects that I already mentioned.
Speaker 2So let's say you're looking ahead. From your perspective, what are the new trends and innovations you foresee in the field of data management and particularly around financial services, say, in the next five years?
Speaker 3So one thing that I'm seeing that is coming up. Number one, the concept of data ops. So data ops, like AI, ops and DevOps and SecOps, data ops become key aspect of the future enterprises and many of them already start adopting it, by the way, but this has become a philosophy of how to manage data within an enterprise and provide a solid data pipeline for different type of use cases. We talk about AI, generative AI, new generation of AI that will be available later. They all really depends on great data pipeline and that's where data ops is going to help you to manage that data pipeline efficiently. Secondly, data observability. That's another aspect of new gen data architecture that needs to be in place to help you to understand what is how your data operate within an enterprise and basically pick up signals that show you there are issues with data flow that needs to be address and that data observability really play an important role on that.
Speaker 3As I said, data, new data architecture, so data mesh, data fabric or fusion of these two something that I've seen is coming because of these two address in different aspect of the data architecture, but that's something that is already been in progress in terms of adaptations or adopting by different enterprises.
Speaker 3Generative AI and AI in general is going to play important role in managing the data not only use data as basically to do to provide different type of insight, but also managing the data ops, helping you manage the data observability, analyze all of those data in real time and automate a lot of these processes.
Speaker 3So this you will see a lot of AI based technology that is going to help dealing with data complexity, data ops and so forth, and that's a new type of technologies that we will see coming up. So you combine the new technologies, combining new basically architecture approach, like data mesh I mentioned before, and new data governance, which is basically includes on-prem and cloud combination, a hybrid architecture that allow you to see your data, understanding how data is being used and providing proof of data usage at any given time, and that's really goes with all of what we talked about compliant with regulations, understanding the security architecture, understanding value of your data and tearing your data security associated with that. And combination of all that is going to be the future trend that we are seeing. We are seeing the start of that already.
Speaker 2And just for like a final question, as we're wrapping up, for a member of the SNEA organisation who is hearing that the current data architecture is no longer sufficient and they may be in agreement with you can see that in their organisation. Where would you tell someone who is looking at that and doesn't know where to start? Who should they be talking to in their organisation and what could they do to just start the conversation in their own org? I know what they can talk about it in SNEA, but in terms of their own org, what would you recommend?
Speaker 3Absolutely so. In some organisation, enterprise architecture teams are good, basically a starting point. In others, definitely, you will talk to also data officers. So we have now AI officers, data officers, compliance officers. It won't be one point of contact, it is a multiple point of contact that you have to take a look at it. So, chief data officers, chief analytics officers, chief AI officer and enterprise data architects and chief compliance officers those are the key stakeholders in actually designing, managing and collaborating of building this modern data architecture. That is really what I will talk to. As I spoke to different institutions. Those are my constituents and stakeholders that I typically talk to.
Speaker 2I appreciate that, parvus. Now how do people stay in touch with you? I know you remember this in the community, but is there any blogs or articles you are sharing? You have a website and so forth where people can follow you.
Speaker 3The easiest way is actually through LinkedIn, so they can reach out to me through LinkedIn. I would be happy to answer questions, and it depends on what else they need to basically to address their issues, challenges and curiosities. I would be happy to do that.
Speaker 2Super. I really appreciate the conversation. I'm sure that we could continue on and on, because it's a big topic, especially when it comes to money. It's a big deal to everyone as we move forward. I really appreciate you coming on with us. Parvus, I would love to talk to you again as you, of course, as things progress, and keep up with you. Thank you for your conversation. Any final thoughts before we end?
Speaker 3I was just going to say one thing remember, when it comes to data, in God, we trust everything else. Everybody else has to show a proof. So that's the fundamental thinking about data how you deal with data. So that's the point that you need to do, but that's about it. I hope everyone enjoyed the conversation that we had and looking forward to future conversations.
Speaker 2Thank you very much. We look forward to you again when we speak to you again in the next episode of Sneha on Data. Thank you, I'm Mark Brown. Thank you very much, parvus. Hope to talk to you again soon. Take care, thanks, mark.
Speaker 1Thank you for listening. For additional information on the material presented in this podcast, be sure and check out our educational library at sniaorg.