The Keys to the Future: AI and Data Analytics with Saket Kumar of Citi

Description

‍

How do AI and data shape the world? And, how can you be the sculptor? In the latest episode of The Curiosity Current: A Market Research Podcast, hosts Stephanie Vance and Matt Mahan welcome Saket Kumar, Vice President of Consumer Analytics at Citi, to explore the true power of deep data analytics and AI in the enhancement of customer experience. Through the episode, they discuss the criticality of AI governance frameworks, balancing automation with human expertise and how organizations can effectively integrate large language models into their analytics workflows. This conversation highlights the intersectionality of generative AI and data analytics, proposing a shift in perception of AI - from a task executor to a thought partner.

Saket leads transformative initiatives in data science and artificial intelligence. With an extensive background spanning aerospace, energy, and finance, he has pioneered the implementation of advanced analytics solutions and cloud computing architectures across multiple industries. His expertise includes leveraging generative AI and large language models to enhance data analytics capabilities and drive organizational innovation. His work in developing conversational analytics solutions and championing AI adoption has helped organizations transition from traditional data analysis to more strategic, insight-driven approaches, making this conversation essential for professionals interested in the future of data analytics and AI implementation.

‍

Transcript

‍

Stephanie Vance: Today, we're delighted to welcome Saket Kumar, VP of consumer analytics at Citi and a trailblazer in the world of data science and analytics. Saket has an impressive background that spans industries like aerospace, energy, and finance with a passion for harnessing advanced analytics and AI to drive innovation and empower teams.

‍

Matt Mahan: From building digital frameworks at Arcadis to leading transformative projects at Citi, Saket has a deep understanding of how technology like generative AI is shaping data analytics and customer experience. Today, we'll dive into his expertise on how large language models like GPT are revolutionizing analytics, bridging enterprise knowledge systems, and driving actionable insights. Saket, welcome.

‍

Saket Kumar: Thank you for your kind words.

‍

Matt Mahan: Of course. We're really happy to have you. So just jumping right in, Saket, your career journey has been pretty extraordinary and unique, spanning a lot of different industries. You've been in aerospace, energy, and finance, all while staying at the forefront of analytics and innovation. What really sparked your interest in data science and analytics? How has your perspective on its potential evolved over the years?

‍

Saket Kumar: Sure. My journey in data science and AI has been driven by my curiosity about how data shapes the world and the power of AI to extract meaningful insights. It all started after my engineering in aerospace, where my career began in data engineering. In 2013, there was a paradigm shift to cloud computing, transitioning traditional database warehousing to scalable cloud architectures for real-time analytics.

I was part of a big aerospace project where we transitioned traditional data warehouses to cloud databases, enabling real-time monitoring of jet engine performance. That experience set the foundation for my career. Later, I moved to the automotive industry, focusing on data analytics and statistical analysis to assess equipment failures and vendor quality. That work helped the industry gauge part performance and increase revenue.

Then, I had the opportunity to work in the energy and environmental sectors in Washington, DC, leveraging AI for water management, transportation planning, traffic optimization, and environmental remediation. These projects had a real impact on improving people's quality of life—what we often call AI for good.

After four years in that space, I transitioned to finance and marketing. Now, I focus on leveraging Gen AI capabilities to transform data analytics and unlock new possibilities through large language models. It's been an exciting journey.

‍

Matt Mahan: That's very cool. In full transparency, I don't consider myself a data scientist at all. So a lot of the questions I have for you may feel like explaining what you do for a living to your parents. Give me an example of how you've used a large language model like GPT to change the way an organization approaches data analytics. What does that intersection look like?

‍

Saket Kumar: The intersection is based on the power of large language models. There was a disruptive shift in 2021 with the release of GPT, followed by ChatGPT. These models can understand questions like humans, orchestrate tasks based on those questions, and generate content.

For organizations, leveraging such powerful models can be resource-intensive. Many fall into the trap of investing without clear revenue-driving applications. The best approach is to use AI as a thought partner rather than just a task executor.

One impactful use case has been conversational reporting. Around 80% of data tasks in organizations revolve around reporting and visualization. Instead of maintaining hundreds of dashboards, AI can generate dynamic reports through conversational UI, translating questions into data queries, retrieving relevant information, and presenting insights in real time. This drastically reduces the time to insights, which is invaluable for leadership decision-making.

‍

Stephanie Vance: That's great. We definitely see that in our business too. It’s such a time saver. I love the idea of AI as a thought partner rather than just an executional arm. Zooming out, how do you see generative AI shifting how businesses approach data analytics? Not just in your role, but across industries. Are there low-hanging fruit capabilities that organizations can easily adopt?

‍

Saket Kumar: Absolutely. The first step in increasing AI adoption is organizational change. Since most employees don’t have AI backgrounds, they need training to effectively use these tools. Organizations should start with AI co-pilot projects, where AI works alongside employees to enhance efficiency rather than replace tasks.

Once employees are comfortable, companies can shift towards product design, incorporating AI more deeply into workflows. Many organizations fail by jumping straight to product design without learning from initial use cases, leading to costly proof-of-concept failures. A structured three-step approach—education, co-pilot adoption, and full integration—leads to better results.

‍

Matt Mahan: That makes sense. In the research world that Stephanie and I sit in, AI is all anyone talks about. Some organizations embrace it, while others fear job displacement. Are you seeing the same range of attitudes in data science?

‍

Saket Kumar: Yes, and some reluctance is actually good. Large organizations often hesitate due to poor data governance. AI models are only as good as the data they learn from. Without structured, high-quality datasets, AI products will struggle to deliver meaningful insights.

Over the years, data analytics has evolved from spreadsheets to business intelligence tools, which forced organizations to improve data governance. With AI, the responsibility to manage data is even more critical. Companies need strong metadata management and structured datasets before effectively leveraging AI.

‍

Stephanie Vance: That makes a lot of sense. Another challenge we hear about is balancing accessibility of insights with accuracy, avoiding oversimplification or misrepresentation. How do you tackle that in a high-stakes industry like finance?

‍

Saket Kumar: There are a few key strategies. One is a hybrid AI approach, where rule-based systems act as checkpoints to prevent AI from generating misleading outputs. Another is explainability—AI should trace and justify its decision-making, allowing users to audit responses.

Other techniques include human-in-the-loop approaches for oversight and retrieval-augmented generation (RAG) to ensure AI stays grounded in verified data. These safeguards help maintain accuracy and prevent AI hallucinations.

‍

Matt Mahan: That’s great advice. Given AI’s rapid evolution, where do you see it heading for data professionals?

‍

Saket Kumar: The future of analytics is conversational. We’re moving towards insight-driven AI interfaces, reducing the need for manual data retrieval. AI co-pilots will revolutionize data workflows, making data management more efficient and cost-effective.

Looking ahead, voice-enabled analytics—where users can talk to their data like Siri or Alexa—is on the horizon. Some companies, like Pyramid Analytics, are already developing such solutions. The goal is real-time, natural language insights available anywhere.

‍

Stephanie Vance: That’s an exciting future. As consumer insights professionals, how can we better collaborate with data science teams to maximize AI’s potential?

‍

Saket Kumar: Strong cross-functional alignment is key. Marketing researchers and business intelligence teams need to understand each other’s workflows. Data science teams must align their analytical lifecycle with business operations. Weekly scrums, open knowledge sharing, and collaborative design thinking will drive better AI adoption and user experiences.

‍

Stephanie Vance: Great answer. Thank you.

‍

Matt Mahan: Absolutely. We appreciate your time and insights.

‍

Saket Kumar: Thank you for the invite. I’ve really enjoyed the conversation. AI is transforming industries, and in a few years, we’ll see even bigger shifts in how businesses generate insights and drive revenue.

‍

Stephanie Vance: For sure. Come back and talk to us again soon!

‍

Saket Kumar: Definitely. Thank you.

‍