Developer Hub

A Python interface for interacting with the Embedded Intelligence Platform (EIP). The EIP uses a proprietary AI architecture called Abstract Generalised Networks (AGNs) to model how complex systems evolve.

Unlike traditional machine learning, AGNs do not require training data, fine-tuning, or retraining. Instead, they autonomously rewrite their underlying algorithms as they observe new data, allowing them to holistically understand how directional shifts influence a system. This approach is ideal for analysing dynamic domains like financial markets and weather patterns, focusing on the structure of change rather than just predicting the next step.

Python Package

To install the sumtyme Python package, you’ll use pip, Python’s standard package manager. Just run the following command in your terminal:

pip install sumtyme

Explainers

AGNs vs. Traditional Architectures

Traditional AI architectures often struggle to model complex, dynamic systems because they’re limited to predicting the next step in a sequence using fixed, linear patterns.

Abstract Generalised Networks (AGNs) use category theory to model the underlying structure of a time series. This provides a holistic view, enabling them to understand the complex forces that drive changes in a system.

AGNs are not autoregressive and analyse the entire data window simultaneously. Therefore, any outputs observed before the last data period of a window should be used for context and not as a prediction. The recommended window size is 5001 data periods.

Information Propagation

Abstract Generalised Networks (AGNs) model a system’s evolution by observing how information propagates across timeframes. This process begins when a directional change on a shorter timeframe creates an early indicator. As the move develops, this indicator propagates to longer timeframes, reinforcing the initial indicator much like a relay system. This propagation continues until the move is fully developed and the indicator dissipates, providing a deeper understanding of a system’s evolution that moves beyond simple next-step prediction.

Demonstrations

• Long Term Directional Change - Visual Example
• Short Term Directional Change - Visual Example
• Intraday Changes - Visual Example

Python Quickstart - Finance

Model Environment

Analyses the environment to forecast a directional change for the final period.

Forecast Change

Forecasts a directional change for the final period.

Rolling Forecast

Performs a rolling analysis to identify evolving trends.