Challenging AI Assumptions in Forecasting: Navigating Human Insight and the ML Conundrum
Forecasting in finance has always stirred debate, with a plethora of perspectives on its methodologies and criteria for accuracy. The emergence of AI and machine learning has only added more complexity to an already complex subject. Can these technologies be relied upon to produce dependable and insightful forecasts? And if so, how can finance professionals optimize their utilization to maximize value?
In this episode of Fintech Corner, JD chats with Yannis Katsaros, Software Engineer at Trovata. Together, they delve into the complexities of uncertainty in forecasting and outline key steps to establish a reliable, balanced approach to leveraging machine learning.
Episode Highlights
- Insights from Yannis’s experiences in data forecasting models across various industries including Finance, Energy, Oil & Gas, and Retail
- The importance of understanding uncertainty at each stage of the forecasting pipeline
- Practical insights for evaluating forecasting models and how this may differ based on the number of data points an organization has access to
- How to effectively compare predicted outcomes with observed data for accuracy and reliability in decision-making
- The critical role of human expertise in validating algorithms for forecasting
Subscribe
Never be out of the loop on what’s going on in finance, tech, and business.
Show Notes / Resources
Read the full transcript.
Learn more about Trovata at https://trovata.io.
About Fintech Corner
A podcast for fintech lovers that starts conversations between bankers, financial technology innovators, and their clients to evolve finance and treasury for the 21st century. For decades, there has been a gap between the technology that banks can offer and what corporate clients need for modern cash flow management and analysis. We’ll cover emerging trends to help bridge this gap, technologies that can reshape the roles of treasury & finance, and the in’s and out’s of API bank connections & open banking.
Yannis Katsaros
Software Engineer, Trovata
Yannis is a Software Engineer at Trovata, primarily working on data and platform engineering. Before working at Trovata, he worked in software engineering, data engineering, and data science roles for companies across various industries including Finance, Energy, Oil & Gas, and Retail. At Trovata he is part of the team responsible for the data platform and API’s that power Trovata’s application, and is focused on leading efforts to continue improving the scalability, reliability, and speed of the platform. Yannis has a Bachelors of Science in Petroleum Engineering from the University of Tulsa, and a Masters of Science in Data Science from Johns Hopkins University.
Joseph Drambarean
CPO, Trovata
Joseph Drambarean is the Chief Product Officer as well as CTO at Trovata. Joseph is one of the founding members of the company and the first engineer. Before Trovata he worked with companies like Capital One where he was at the forefront of digital transformation, leading product management as head of the innovation labs and mobile banking teams. Joseph is driving innovation around rapid deployment and customer onboarding, bank-grade security, and machine learning at Trovata, creating a more consumerized user experience for customers from small businesses to enterprises.
