As the world’s largest quant finance event, QuantMinds International 2023, has officially drawn to a close we reflect on this fruitful and well organized conference, which undeniably exceeded expectations. The debut of the joint hackathon, coupled with both insightful AI-themed presentations by globally recognized speakers and practical workshops, contributed to the resounding success of this year’s QuantMinds conference.
QuantMinds and CompatibL partnered this year to deliver their first ever annual hackathon, held as part of QuantMinds International 2023. Within the framework of this hackathon, which took place at the conference on November 13, participants were able to showcase the capabilities of large language models (LLMs) in quantitative finance through the adept use of coding or prompt engineering.
Alexander Sokol, CompatibL’s Executive Chairman and Hackathon leader, said:
“We wanted to help people to see for themselves that these tools work, and to gain first-hand experience using them. We hope events like this will accelerate the adoption of this game-changing new technology and help us as financial professionals to better serve our clients as well as society at large.”
This year the competition consisted of two streams, with each stream further subdivided by model type (GPT or LLAMA), giving a total of four categories:
- Pricing Model Source Code Comprehension Stream (Source-GPT and Source-LLAMA categories)
- Trade Confirmation Comprehension Stream (Text-GPT and Text-LLAMA categories)
Participants could choose between natural language and Python code submissions in each of the four categories.
The hackathon proved to be a definite success, gaining widespread recognition from both participants and organizers. The teamwork by participants, coupled with the careful planning and execution by the organizers, played a key role in the event’s effectiveness in promoting innovation and collaboration, successfully illustrating the use of LLMs in quant finance and allowing winners to test their coding abilities and gain skills they could use to do their jobs better and more quickly.
The Hackathon Winners
We congratulate all the hackathon winners and wish them continued success in their endeavors! Their outstanding performance and collaborative mindset rightfully earned them a place on the leaderboard and some valuable prizes.
The hackathon leaderboard is available here.
Alexander Sokol’s Presentation and Workshop
Aligned with contemporary trends, the CompatibL team recently developed a cutting-edge AI solution to address model governance and document comprehension in quant finance using LLMs.
Iin his QuantMinds conference presentation, “Customizing Large Language Models for Quant Finance Applications,” Alexander Sokol highlighted the rationale behind this solution. He also hosted a workshop on “Large Language Models in Financial Markets,” diving deeper into the AI technology.
“In our industry, a tremendous amount of highly specialized documentation must be produced, analyzed, and kept in sync with the evolving models and technologies. CompatibL AI acts as a co-pilot for model governance experts, and it can accelerate certain tasks by a factor of 10 or more.”
There was significant interest both in the LLMs for model governance and document comprehension and in the specific approach we have employed to address this essential banking function. The concept garnered widespread attention, as did the nuanced details of our application of LLMs in this context.