This year has been transformative for the data and analytics industry. Data leaders today are racing alongside the rapid advancements in technologies like generative AI, balancing cautious optimism with the need for fast adoption.
As we approach 2024, we brought together the following financial services experts in a roundtable discussion – “Enterprise Data Leaders: Opportunities & Predictions for 2024,” to share reflections on 2023 through buzzwords and explored opportunities, and predictions for 2024. They also dissected the strategic approaches to AI innovation.
Satya Choudary, Credit Suisse VP, Data & Analytics - Global Head of Cloud Platform, Data Engineering & Management, Machine Learning and AI
Sudip Ghose, Equifax Vice President, Enterprise Data & Analytics
Alex Tait, BMO US Chief Data & Analytics Officer, Enterprise Data & Analytics
Jim Tyo, Invesco Chief Data Officer
The conversation was moderated by Jenna Boller, Ocient Senior Director of Marketing.
We will highlight the key takeaways from the highly insightful discussion and explore the complexities of data ecosystems, talent acquisition, and the transformative potential of emerging technologies.
Buzzwords often reflect emerging trends, technologies, or concepts that shape discussions and innovations around us. Here are the top buzzwords for 2023 shared by the panelists:
Invesco’s Tyo points out “data observability.” He presents it in the backdrop of the turmoil experienced by the financial services industry, including market fluctuations that led to rethinking strategies and being business-focused.
Tait from BMO picks “anxiety.” This is in light of economic constraints, ongoing wars, and the challenges, opportunities, and skills needed for generative AI.
It is “GenAI” for Ghose from Equifax. Commenting on a productivity-led use case of generative AI, he states that machines can now write better letters than humans and can enhance customer experience.
Choudary from Credit Suisse chooses “sustainability.” He believes that both data and AI have the power to enhance sustainability, whether it is climate control, agriculture, or supply chain.
Kickstarting the conversation, Boller lays down the following predictions for 2024:
As AI grabs headlines, data quality will remain a critical barrier to tech enabled enterprises.
More industries will analyze climate intelligence data.
More enterprises will architect for hyperscale.
Sharing his opinion on the forecast, Tyo says that Environmental, Social, and Governance (ESG) is a key aspect especially in the asset management industry. Choudary adds that climate control makes a lot of sense around sustainable investing: “The first prediction on data quality is paramount because garbage in, garbage out. Without the right data in place, there will be no unbiased AI predictions.”
Tait equates the situation with the game of Jenga which requires a solid foundation. He maintains that the next generation of data practitioners are more interested in data science aspects of the role than the management aspect which requires conscious attention.
The panelists add the following top 10 predictions to the list:
There will be an increase in issues around security and quantum computing, which is also tied to climate control and sustainability. There will also be a need for enhanced and efficient computing with AI scaling. — Choudary
There will be advances in assistive technology, and there must be a human in the loop. Assistive technologies that aid productivity will creep into cars, phones, and watches. — Tait
There will be an increased need for transparency with a better understanding of risks. This will drive all industries to standardized disclosure. — Tait
Organizations will see a lot of experimentation and brainstorming to deliver better with new technologies. Centers of excellence (CoEs) must be put together with all the sectors working together to deliver what is needed. — Ghose
There will be AI governance topics and, like in data governance, there will be a conflict of interest. — Ghose.
Security threats will compete and continue to evolve. — Choudary
Resource shortages will continue in computing and good talent. — Choudary
There will be a race between the overall challenges of climate control. — Choudary
Every piece of capital invested needs to be a business priority. All foundational components built up in data organizations or ecosystems will be tied directly to business value. — Tyo
There will be no data strategy but business strategy in a data-driven world. — Tyo
AI has taken the world by storm. While it offers the opportunity to experiment, there might be expensive failures, too. Sharing his approach to riding the AI wave, Ghose says that putting together a team that knows the company goals is critical. He states that data goes deep into the organization and can assess its performance, and once a solid foundation is laid, the next layer is of innovation.
Understanding how to monetize the data and create new products for the company in the current scenario is critical, says Ghose. “Organizations must focus on investing right and getting productivity out in 2024.”
Tait recommends leveraging AI in client channels to help with client relationships. Being in the heavily regulated financial services industry, he urges doing it thoughtfully with guardrails in place.
“The challenge lies in understanding the risks and how to mitigate them, as governance cannot take a backseat, whether it is data or AI governance,” says Tait. However, he sees opportunities and reveals that his organization is starting slow with internal-facing use cases for Gen AI with a measured, thoughtful approach.
On a similar note, Tyo believes there will be an exponential climb in the maturity and the impact that could be seen from some of the new and upcoming principles and practices. Echoing Tait, he mentions starting small with an internal model and balancing innovation with risk.
Digital twins, data fabric, and data mesh continue to be prominent topics within the tech discourse.
Choudary emphasizes that the choice of strategies hinges on specific business cases. “The focus should be to look for an optimal path to solve the problem, whether it is data mesh, a cloud solution, generative AI, or a small solution built in Excel.”
Choudary encourages exploring data mesh and fabric for strategic reasons. While data mesh provides flexibility to different business units and domains, data fabric is used to consolidate with a centralized view of the data, and leverage its power. “However, everything must tie back to problem solving,” he says.
Data mesh and data fabric are “favorite deliverable concepts” for Tyo. He mentions building a fabric across Invesco. It is driven by a business case around taking advantage of the investments made previously while being able to invest in new consumption layers or solutions within the ecosystem. The next step is to create a common experience for the end consumer and business partners.
Tyo adds that fabric is built as part of the ecosystem, and mesh is a part of the process of delivering data products efficiently for business.
Ghose describes mesh as a more modern concept born out of the emergence of data platforms and virtualization techniques. “It aims to address both the cost and time value associated with data, and make it more efficient to get the data in and to the consumer.”
The evolving landscape of data ecosystems is accompanied by a noticeable increase in complexity, prompting a deeper exploration of the contributing factors.
Ghose mentions the proliferation of systems and how competing hyperscalers play a role in making it sound more complicated. To resolve this, he urges CDOs and CTOs “to engage with consultants, partner up, and build a roadmap.”
Emphasizing the cost aspect, Ghose states that organizations must try and consolidate things to make it cost-effective and efficient.
In the same vein, Choudary points out two key contributing factors to the overcomplication. First, a rise in data consumption for gaining business insights. Second, increased mergers and acquisitions in specific industries, indicating overall business evolution.
According to Tyo, if an organization is not digitally native and has undergone merger and acquisition activity, it will have a complex ecosystem. To address this, Invesco is going through a technology simplification journey. “With business moving fast, things will get complicated but they should not affect the client,” Tyo adds.
Shortage of skilled data talent has been a continuing issue in the data space. The advent of generative AI has renewed the demand along with the need for new skills like prompt engineering.
Sharing his approach to staffing, Choudary affirms looking mostly for data analysts who have holistic knowledge across data engineering, governance, and AI, and can also upskill into prompt engineering.
Following up, Tait encourages organizations to have a strong value proposition and differentiate themselves as employers.
Tait adds that COVID has accelerated virtual work, allowing access to talent pools from unusual places. Chiming in, Choudary mentions establishing partnerships with local universities to attract talent.
The talent challenge is more on the cloud engineering and data cloud engineering side for Ghose. He says, “It is harder to retrain older resources than hire freshers, and train them in the mindset and methodologies.”
Tyo puts emphasis on being as lean as possible, especially being the cost center, not a revenue-generating organization. The two factors that attract more diverse talent at Invesco are:
Global nature that leads to diverse talent growth
Reskilling and upskilling
“Bringing together those two pieces will help the organization try to navigate the talent scarcity and the competitiveness in the data analytics space,” says Tyo.
Concluding the discussion, Tait foresees an arms race between companies with the current trajectory of AI. Sharing an example, he mentions how client-centric differentiation backed by automation improved customer satisfaction for airline companies.
Sharing his thoughts, Choudary says that with the evolving technologies, organizations must take a step back and understand how to adopt a technology. He recommends having a holistic approach and mentions a four-step framework:
Ensuring executive buy-in on key adoptions
Building the culture with right people and upskilling/reskilling them
Having the right governance
Having the right structure in place
Adding his view, Ghose urges data leaders to consider buying versus building. For him, the ideal way is to train humans to build innovative products and then buy the products to serve business value. “Understanding what to build and what to buy will lead leaders to success,” says Ghose.