Opportunities and challenges for enterprise AI: Deep Ratna Srivastav interview

Opportunities and Challenges for Enterprise AI: Deep Ratna Srivastav Interview

AI is more than just exciting. With the right strategy and implementation tied to specific business outcomes, AI can be very practical. Personalized investment portfolio management, next-gen interactive dashboards, and AI-powered customer support are just the beginning.

This was the topic of conversation between Nazar Labunets, product marketing manager at Ataccama, and Deep Ratna Srivastav, SVP Head of AI and Digital Transformation at Franklin Templeton.

Franklin Templeton is a global leader in asset management. Deep heads a team that builds AI models to provide more accurate and personalized financial advice. Recently, he's worked on projects like a goals optimization engine and support functionalities to deliver greater information capture and more customizable capabilities to the company and its clients.

Watch their whole conversation and read the key insights below.

AI strategy, opportunities, and creating demand internally

Deep begins the interview by covering the three key dimensions of his position at Franklin Templeton. Then, they discuss the need to develop internal demand for AI capabilities, Deep's AI strategy, and value creation.

Key takeaways:

  • The role of Head of AI comes down to three key dimensions:
    • Creating demand. Work with business leaders and their teams to create a vision of AI's potential, then use it to create demand internally and externally for those capabilities. Build a realistic long and short-term vision.
    • Deliver on promises. AI promises must be fulfilled to obtain resources and support for projects. This takes a lot of collaboration between internal and external parties. Technology vendors, product management, research & development, and SMEs must come together to deliver on the objectives.
    • Business transformation. As you deliver on new capabilities, determine how they impact the business and its value proposition, and tangibly measure their success.
  • Building an AI strategy around the outcomes impact on the business. How can it transform the enterprise, the ecosystem, and the way it behaves, and how will you get there? Piecing together the answers to all these questions will result in an intensive roadmap that will ultimately become your AI strategy.
  • How to turn AI into magic? Design models for specific tasks trained individually. Once you get them right, begin interconnecting them, bringing significant and positive change to the whole enterprise.

Generative AI, the importance of data quality and governance, and essential data management capabilities for AI projects

For the next stage, Deep and Nazar discussed easing the reality of gen AI, why AI models need data quality and governance processes, and the best ways to ensure high-quality data.

Key takeaways:

  • Gen AI is just a user-friendly interface for AI capabilities that have already been here for a long time. Still, Gen AI made them accessible to a whole new group of people with different skill sets, breaking down data silos in a meaningful way.
  • However, that same interconnectedness between machines that breaks down silos amplifies the potential damage in case of an error. This makes data quality and governance crucial for AI.
  • Explainability is key to any successful AI program. It was easier to see where decisions came from in the past because they were human-oriented. Organizations need insight into how AI systems work and why changes happen to trust their results.
  • Deep listed the following essential technologies for managing AI model training data:
    • Stress/simulation/scenario testing.
    • Synthetic data creation.

Challenges with AI, delivering on ROI, creating repeatable processes, and AI tips for beginners

In the final part of the interview, Nazar and Deep cover the common challenges companies face when adopting AI, how to deliver immediate results to validate investments, the importance of repeatable processes, and tips on getting started with AI for beginners.

  • Bridging the gap between technical and business teams requires a huge shift in talent. To break down silos, you need talent which can cut across boundaries, spread understanding across all teams, and create a dialogue.
  • Determine your company's priorities, evaluating which capabilities and parameters are important to them. Then, use those priorities to formulate your ROI plan, first delivering what your company needs most.
  • Employ product thinking to develop repeatable processes. Be sure to synchronize the entire effort across all relevant teams.
  • Tips for AI beginners:
    • Start small with day-to-day decisions, staying as specific as possible.
    • Pick intuitive tasks.
    • Choose capabilities that will further your understanding of AI.
    • Use this framework and experience to develop more complex processes on larger volumes of data.

Key Takeaway: The era of opportunity is upon us

Deep concluded his interview with a wave of optimism. He thinks we're entering unprecedented times regarding AI and its potential with generational opportunities at our doorstep. His advice on seizing this momentous opportunity is simple: connect, keep the energy going, and be aware that we are in a fortunate space and time – let's do it right.

Want to learn more about becoming a successful CDO? Visit the CDO Success Hub to learn about different opportunities and technologies for prospering in your position.

Arrow Right
Arrow Left