AIstack Features

AIstack is a low-code deep learning solution. But how does it work?

Type-Based Abstraction

Type-Based Abstraction

AIstack currently supports twelve data types (Text, Category, Numerical, Binary, Sequence, Set, Image, Time Series, Audio, H3, Date, and Filter Bank). New data types (Videos, Graphs) are coming soon.

Encoder-Combiner-Decoder (ECD) Architecture

Encoder-Combiner-Decoder (ECD) Architecture

A general modularized deep learning architecture called Encoder-Combiner-Decoder (ECD) that can be instantiated to perform a vast amount of machine learning tasks. Every AIstack model is defined in terms of:

  • encoders that encode different features of an input data point,
  • combiners that combine information coming from the different encoders, and
  • decoders that decode the information from combiners into one or more output features.

Declarative Model Definition

Declarative Model Definition

Declarative model definition configuration files enable inexperienced users to obtain effective models and increase the productivity of expert users.

Ready to start?

A member of our team will walk you through the platform and demonstrate how our solution can help!