AI Investments uses advanced AI methods for developing robust investments strategies

AI Investments Ltd is a fintech start-up that was founded to create an innovative platform for optimizing financial portfolio management using the most advanced AI and deep learning solutions available on the market.
Our goal is to create the first, fully automated, AI-based trading solution. We are implementing different trading strategies and support them with AI methods.
We are using AI for financial time series prediction and risk control/position size management.
AI Investments Ltd was established in 2018 by an experienced manager and investor Dominik Libicki (co-founder of Insignis TFI) and successful trader and IT system architect Pawel Skrzypek.
The Luxembourg registered Omphalos Fund started to use AI Investments technology for investments portfolio management in November of 2021.AII solution is offered as a Software as a Service platform for investment companies, hedge funds and proprietary trading companies. It could be used in two models:
With AI Investments strategies, customized and parametrized for given customers in terms of risk, position-sizing, frequency of trading.
As a generic assets pricing forecasting to support customer’s investment strategies with AI capabilities.AI Investments utilizes the latest scientific results in the area of artificial intelligence to build their products. In particular advanced deep learning architectures customised for time-series prediction are used for building predictive models for financial instruments [1,2]. These predictions are then fed to the trading  strategies developed with the help of reinforcement learning algorithms [3]. Deeper understanding of the algorithm’s decisions is obtained using the state of the art algorithms for analysing the influence of input factors on the predictions [4].
  1. Lim, B., Arık, S. Ö., Loeff, N., & Pfister, T. (2021). Temporal fusion transformers for interpretable multi-horizon time series forecasting. International Journal of Forecasting.
  2. Smyl, S. (2020). A hybrid method of exponential smoothing and recurrent neural networks for time series forecasting. International Journal of Forecasting36(1), 75-85.
  3. Singh, Satinder, Andy Okun, and Andrew Jackson. “Learning to play Go from scratch.” Nature 550.7676 (2017): 336-337..
  4. K. Mnich, W.R. Rudnicki. All-relevant feature selection using multidimensional filters with exhaustive search. Information Sciences 524 (2020), 277-297.