In this Leading with Data episode, Eleni Verteouri, AI Tech Lead and Director at UBS, shares her invaluable insights on the transformative journey of AI in finance. With over a decade of experience in model development and a prestigious Forbes Cyprus 20 Women in Tech Award 2024 recognition, Eleni has been at the forefront of shaping modern financial technologies. In this podcast, we delve into her expertise, exploring the evolution of AI, from traditional mathematical models to innovative solutions. Join us as we uncover the key insights and learn from Eleni’s remarkable career.
You can listen to this episode of Leading with Data on popular platforms like Spotify, Google Podcasts, and Apple. Pick your favorite to enjoy the insightful content!
Let’s look into the details of our conversation with Eleni Verteouri!
I’ve always leaned towards technical subjects, but it wasn’t clear what direction to take. During my first engineering degree, I explored GPUs and distributed computing. At that time, cloud computing wasn’t as prevalent, and finance seemed to demand high computational power for complex calculations. This curiosity led me to financial mathematics, which eventually took me to Switzerland and into the banking sector. It was a pivotal moment in my journey into technology and AI.
After completing my engineering degree with a focus on computers, I wanted to delve deeper into modeling. I pursued quantitative finance, a blend of mathematics and finance, at ETH Zurich and the University of Zurich. This education, combined with my master thesis at an investment bank, paved the way for my internship and subsequent roles in the industry.
AI has transformed the banking industry significantly. Initially, the focus was on traditional financial theories and mathematical constraints. Over time, AI has been integrated into personalized services, prototype building by end-users, and risk estimation. It’s been a journey from operating within a strict framework to adopting AI for efficiency and innovation.
Thin data problems were traditionally tackled by individuals using statistical models and rule-based approaches. Today, we have a more robust data infrastructure and governance, and generative models that don’t necessarily require historical data. They can generate synthetic data and assist with metadata, offering new angles to address these challenges.
My role involves working on an internal generative AI platform designed to support various use cases, integrate with other tools, and foster internal collaboration. My focus is on safety and security, ensuring responsible AI and user-friendly data science practices. I also contribute to setting development guidelines and fostering academic partnerships.
Education is crucial, both for users and developers, as roles are blending with engineers working with AI and data scientists driving adoption. Discussions about what “responsible” means are vital, as are partnerships with startups filling gaps in safe and responsible AI adoption.
We’ll see a rethinking of business and product perspectives, governance, and how we work. Generative AI, considerations around artificial general intelligence, and the integration of AI into personal life are key trends. Safety, security, and sustainability will remain paramount.
Ensuring that AI innovation benefits a fair portion of the population is a challenge. It’s about making sure that new knowledge and skills are accessible and attainable for most people, which may require more coordination and communication.
With traditional machine learning, it was easier to identify and control biases. LLMs present a challenge in explainability. Agents could be part of the solution, helping simulate scenarios and providing a bandwidth of safe and acceptable operation. Human oversight remains an essential part of the process.
I’m excited to see problems traditionally solved by AI being addressed with LLMs, such as document parsing and insight generation. I’m also interested in AI’s potential to improve mental health, provide empathy, and help neurodivergent individuals integrate better into society.
As a model developer and AI product manager, I find the boundaries between manual work and AI assistance are blurring. I use AI for coding, troubleshooting, proofreading, and brainstorming. It has made me more independent and efficient in building solutions.
Empowerment through technology is an equalizer. It’s important to judge individuals based on capabilities rather than appearances. We should ensure inclusivity, provide equal opportunities, and encourage women to be outspoken, confident, and assertive in their careers.
Eleni Verteouri’s journey and expertise offer a unique perspective on the future of AI in finance. Her contributions, marked by her Forbes award and academic achievements, provide a roadmap for the industry. As we navigate the complexities of AI, Eleni’s emphasis on responsible practices, education, and ethical considerations serves as a guiding light. Stay tuned to Leading with Data for more inspiring sessions and insights into the world of AI and data science.
For more engaging sessions on AI, data science, and GenAI, stay tuned with us on Leading with Data.