Prism
Prism by Borealis AI is a simulated financial arena that supports algorithmic trading on a fictional stock called “PRSM.” Agents submit trade orders via an API with the aim of forecasting and exploiting one another’s actions, testing and perfecting their trading strategies and algorithms. This artificial exchange was built to help expose those interested in AI and finance to AI-based trading.
The Platform
-
The Prism platform evaluates agents’ models, offers order book visibility, and gives agents the ability to execute trades. Additionally, agents can utilize platform metrics to compare their model’s performance with that of their competitors.
-
A competition episode consists of a default trading agent, the Fundamental Agent (FT), which provides market liquidity and default market dynamics. The FT’s trading strategy remains undisclosed and varies between episodes.
-
To create their trading agents, participants can either write scripts using the Prism API schema or our Starter Kit. The platform offers a testing environment to ensure agents can interact with the Prism server.
-
The server provides access to a dashboard displaying market info such as recent trades, bid/ask data, and summary statistics. It also features a leaderboard that ranks participants based on their realized profits.
Your Platform to play.
Experience the competition
Borealis AI hosts its first algorithmic trading competition. Take a look at the Prism competition in action.
The Prism Trading Competition
We invited experts in forecasting, trading, and adversarial attacks to test their models on Prism to showcase how AI can be used in finance, increase financial literacy, and provide a safe and hands-on learning opportunity for the wider ML community.
At Borealis AI, we have a mission to create real-world impact through scientific pursuit; Prism competition reflects our belief that exploration – while at times challenging, can also be fun.
Anchored in
Research
In response to a need for a competitive environment to test the performance of time-series algorithms, the Engineers at Borealis AI created Prism, a gamified, simulated environment that mimicked the behaviour of a competitive market.
Prism offered Researchers a platform to evaluate their time-series algorithms against competing agents to see how their models stack up. Additionally, we intend to use Prism to support our North Star research on Non-Cooperative Learning in Competitive Environments.
-
Agent Forecasting at Flexible Horizons using ODE Flows
Agent Forecasting at Flexible Horizons using ODE Flows
A. Radovic, J. He, J. Ramanan, M. Brubaker, and A. Lehrmann. International Conference on Machine Learning Workshop on Invertible Neural Nets and Normalizing Flows (ICML), 2021
-
Efficient CDF Approximations for Normalizing Flows
Efficient CDF Approximations for Normalizing Flows
C.S. Sastry, A. Lehrmann, M. Brubaker, and A. Radovic. Transactions on Machine Learning Research (TMLR), 2022
-
Continuous Latent Process Flows
Continuous Latent Process Flows
R. Deng, M. Brubaker, G. Mori, and A. Lehrmann. Conference on Neural Information Processing Systems (NeurIPS), 2021
-
Generating Videos of Zero-Shot Compositions of Actions and Objects
Generating Videos of Zero-Shot Compositions of Actions and Objects
M. Nawhal, M. Zhai, A. Lehrmann, L. Sigal, and G. Mori. The European Conference on Computer Vision (ECCV), 2019
👋 Meet the team behind the platform
Engineering plays a vital role at Borealis AI, turning our machine learning models into products that generate exponential business value. We understand the financial system we operate in, and we have the resources, knowledge and tools to improve it.
Our engineersBlog