This year, QuantMinds International, the world’s largest conference for quantitative finance professionals, once again brought together industry leaders, academics, and technology innovators to discuss cutting-edge topics and emerging trends in finance, including artificial intelligence/machine learning, large language models (LLMs), risk management, and quantitative modeling.
CompatibL highlighted its innovative approach in three unique sessions by Alexander Sokol, the company’s Chairman and Head of Quant Research.
Alexander ran a workshop on “Practical techniques for building reliable LLM-based workflows,” and gave a joint presentation on “Autoencoding term-structure models” with Andrei Lyashenko and Fabio Mercurio. He also gave a keynote presentation on “What-if trade entry from email using LLMs.”
The latter covered CompatibL’s TradeEntry AI solution, which is designed to automate the trade entry process using LLMs. Unlike traditional software, LLMs behave more like humans—that is, they excel at interpreting data but can also misinterpret details or produce inconsistent results. To address this, CompatibL developed workflows that replicate human strategies for ensuring reliability.
QuantMinds-CompatibL Hackathon
In partnership with QuantMinds, CompatibL also held its second annual hackathon at the conference, uniting some of the brightest minds in quantitative finance and AI to tackle the task of “Trade entry for what-if analysis.” The participants had the option to compete using prompt engineering without coding or to submit coding in Python.
Winners
- The XP team won in both the Coding GPT and Coding Llama categories, and the overall Grand Prize
- The BofA team won in the non-coding Llama category
- Andrea Frecassetti from UBS won in the non-coding GPT category
Congratulations to all the winners!
QuantMinds International once again proved to be a major networking hub, fostering dialogue and innovation among quants worldwide.
