Video Type: Conference

Watch the recordings from the 2025 Quant Insights Conference.

Video Description

To ensure and monitor large language models (LLMs) reliably, various evaluation metrics have been proposed in the literature. However, there is little research on prescribing a methodology to identify a robust threshold on these metrics even though there are many serious implications of an incorrect choice of the thresholds during deployment of the LLMs. Translating the traditional model risk management (MRM) guidelines within regulated industries such as the financial industry, in this talk, I will discuss a step-by-step recipe for picking a threshold for a given LLM evaluation metric. I will emphasize that such a methodology should start with identifying the risks of the LLM application under consideration and risk tolerance of the stakeholders. We then propose concrete and statistically rigorous procedures to determine a threshold for the given LLM evaluation metric using available ground-truth data.

Speaker Bio

Dr. Dhagash Mehta

Dr. Mehta is the Head of Applied AI and Responsible AI Research (Investment Management) at Blackrock Inc. and an Editorial Board Member at the Journal of Financial Data Science and Journal of ESG and Impact Investing. He was also General Chair for the ACM International Conference on AI in Finance 2024 (ICAIF24). Previously he was a Senior Manager,Investment Strategist (Machine Learning – Asset Allocation) at The Vanguard Group, a Senior Research Scientist at United Technologies (UTX) Research Center, a Research Professor at University of Notre Dame, a Fields Institute Postdoc Fellow, a Visiting Fellow at Simons Institute for Theory of Computing at Berkeley, a postdoctoral researcher at the University of Cambridge, Syracuse University (USA) and National University of Ireland Maynooth (Ireland) and a research student at Imperial College London (the UK), the University of Adelaide (Australia).

Video Description

Commodity futures markets are notoriously difficult to model due to their sparse disclosures, idiosyncratic fundamentals, and the complex web of macroeconomic, geopolitical, and physical flow drivers. In this presentation, I introduce a novel research framework that leverages Knowledge Graph (KG) triples and Graph Neural Networks (GNNs) to enhance the modeling of cross-sectional and temporal dependencies in commodity futures.

We begin by constructing a domain-specific knowledge graph from structured and unstructured data, using large language models (LLMs) as semantic extraction tools. Document sources include news articles, regulatory filings, and macroeconomic event reports. This text corpus is transformed into temporal entity-relation triples that encode time-aware relationships between entities.

Each node in the graph represents a commodity future, optionally enriched with standard price-derived features. Edges capture latent relationships, both explicit and inferred, such as shared production dependencies or co-movement during stress regimes. By applying GNN architectures, we learn context-aware embeddings for each futures contract, which are then used for return forecasting and relative value positioning within a long-short investment framework.

Speaker Bio

Tony Guida

Tony Guida is a Quantitative Portfolio Manager, researcher, and author with expertise in systematic investing, machine learning, and financial data science. He currently works on the integration of large language models (LLMs) and knowledge graphs into investment processes for a Swiss asset manager specializing in growth-oriented thematic equity portfolios.

Tony began his career at Unigestion in 2006, joining the quantitative equity team as a research analyst. He later became a member of the Research and Investment Committee for Minimum Variance Strategies, where he led the factor investing research group for institutional clients. In 2015, he joined EDHEC-Risk Scientific Beta as a Senior Consultant, focusing on risk allocation and multi-factor strategies. In 2016, Tony moved to a major UK pension fund, where he built an in-house systematic equity strategy and co-managed a £8 billion portfolio as Senior Quantitative Portfolio Manager.

In 2019, he joined RAM Active Investments as a Senior Quantitative Researcher in equities, later co-heading the systematic macro hedge fund offering. In 2023, Tony co-founded a hedge fund, where he served as Co-Head of Research and Portfolio Manager, leading the development of cross-asset, machine learning-driven investment strategies.

Tony holds Bachelor’s and Master’s degrees in Econometrics, and a Master’s in Economics and Finance. He has authored and edited several industry-leading books, including:

• Big Data and Machine Learning in Quantitative Investment (Wiley, 2018)
• Machine Learning for Factor Investing: R Version (CRC, 2020)
• Machine Learning for Factor Investing: Python Version (CRC, 2023)

He is also an Advisory Board Member for the Financial Data Professional Institute and serves as a peer reviewer for academic journals in machine learning and finance.

Video Description

Artificial Intelligence (AI) and Machine Learning (ML) are having a profound impact on systematic investing and trading, fundamentally transforming quantitative strategy development and execution. This talk highlights recent advances in applying AI and ML to systematic investment management, emphasizing their practical roles in generating alpha, optimizing portfolios, and enhancing trading strategies. We argue that maintaining a human-in-the-loop is essential to effectively address challenges and fully realize the benefits of ML-driven automation. The presentation concludes by highlighting promising directions, opportunities, and challenges ahead.

Speaker Bio

Dr. Petter Kolm

Petter Kolm is a Clinical Professor at NYU’s Courant Institute, where he directs the Mathematics in Finance Master’s program. In 2021, he was honored as “Quant of the Year” by Portfolio Management Research (PMR) and the Journal of Portfolio Management (JPM) for his significant contributions to quantitative portfolio theory. He has co-authored numerous influential articles and books on quantitative finance and financial data science. Petter serves on several company advisory boards (including Aisot, Axyon AI, and GoQuant), editorial boards for leading academic journals, and boards of directors for professional associations. Previously, he worked in Quantitative Strategies at Goldman Sachs Asset Management. As a consultant and expert witness, Petter provides expertise in machine learning, portfolio management, risk management, and systematic trading. He earned his Ph.D. in Mathematics from Yale University, an M.Phil. in Applied Mathematics from the Royal Institute of Technology (KTH), and an M.S. in Mathematics from ETH Zurich.

Video Description

Autoencoders can be used to accurately describe the shape of yield curves by a low number of factors, typically 2-3. We demonstrate this using cross sectional data across 10 different currencies spanning a period of 12 years. Using dynamic arbitrage arguments we show that the risk-neutral dynamics of the yield curves can be deduced from the possible shapes produced by the auto-encoder. We use this come up with a new decomposition of fixed income portfolio returns. We draw connections to Cheyette models and discuss application to other asset classes.

Speaker Bio

Dr. Jesper Andreasen

Jesper Andreasen is head of quantitative analytics at Verition Fund Management. Jesper’s career spans 28 years in the industry and he has previously headed quant teams at Saxo Bank, Danske Bank, Bank of America, Nordea and General Re Financial Products. Jesper co-received Risk Magazine’s quant of the year awards in 2001 and 2012 and their in-house risk system of the year in 2015.

He is an honorary professor of mathematical finance at the University of Copenhagen and he holds a PhD in the same subject from Aarhus University.

Video Description

Agentic AI is gaining visibility in finance, but many so-called “agentic systems” are little more than PowerPoint proofs of concept or API wrappers with no real autonomy. This talk goes beyond the hype and explains how real agentic systems are being built and deployed. From tool-augmented reasoning and structured reflection loops to infrastructure-aware orchestration. The session outlines what it takes to move from concept to production – cutting through agent-washing and showing where tangible value is already being created.

Speaker Bio

Nicole Königstein

Nicole is the Co-Founder, CEO, and Co-Chief AI Officer at Quantmate, a deep-tech fintech company developing AI agents for portfolio management and strategy development via natural language. She is a globally recognized thought leader in large language models (LLMs) and agentic architectures, with a particular focus on their transformative applications in quantitative finance.

As a guest lecturer, Nicole shares her expertise in Python, machine learning, and deep learning at universities. She is also a frequent speaker at AI and quantitative finance events. Nicole has authored Math for Machine Learning and Transformers in Action with Manning Publications. Her forthcoming book, Transformers: The Definitive Guide – Applications Beyond NLP, will be published by O’Reilly Media.

Video Description

AI and humans perform surprisingly alike in classic behavioral psychology experiments. These experiments show that AI shares many cognitive biases of humans and is prone to human-like errors of logical reasoning and recall. We will examine the reasons for this surprising and perhaps even shocking finding, some superficial and some profound. Our analysis will lead to concrete, practical techniques for avoiding these biases and increasing the reliability of AI in business settings.

Speaker Bio

Dr. Alexander Sokol

Alexander Sokol is the founder, Executive Chairman, and Head of Quant Research at CompatibL, a trading and risk technology company. He is also a co-founder of Numerix, where he served as CTO from 1996 to 2003.

Alexander won the 2018 Quant of the Year Award together with Leif Andersen and Michael Pykhtin for their joint work revealing the true scale of the settlement gap risk that remains in the presence of initial margin. Alexander’s other notable research contributions include systemic wrong-way risk (with Michael Pykhtin), joint measure models and the local price of risk (with John Hull and Alan White), the use of autoencoder manifolds for interest rate modelling (with Andrei Lyashenko and Fabio Mercurio), and the mean reversion skew.

Alexander graduated from high school at the age of 14 and earned a PhD from the L.D. Landau Institute for Theoretical Physics at the age of 22. He was the winner of the USSR Academy of Sciences Medal for Best Student Research of the Year in 1988.