With the Ekkono SDK you can bring intelligence into your IoT devices. We currently have 4 different products in our portfolio. This is a short product overview of our SDK. Feel free to contact us for more information.
- Ekkono Primer (C++) – Creates and exports machine learning models.
- Ekkono Edge (C++) – Model inference and training. Flash requirement approx 1MB.
- Ekkono Crystal (C) – Model inference and training. Flash requirement approx 30 kB. Platforms without OS.
- Ekkono Spectral (C) – Streaming signal processing.
Primer contains Python bindings for all libraries for easier model development. In effect, you will usually construct the models in a Python environment. You may also pretrain the models on collected data in Python, but our online models will really excel once they reach the individual devices.
With the developer license, you will get access to the following resources:
- Libraries prebuilt for desktop use (for development in Python).
- Libraries delivered as source (for deployment on your target platform).
- Ekkono Studio – Online development environment with pre-installed libraries.
- Interactive machine learning tutorials in Jupyter Notebooks.
Ekkono SDK Functionality
- Removal of outliers and missing values
- Filtering of columns and rows
- Feature elimination
Preprocessing pipeline for streaming data
- Lags, moving averages, expressions
- Supports incremental training
Incremental and Batch Modeling
- Neural Networks, Ensembles, Random Forest, Decision Trees, Linear Regression
- Anomaly and Concept Drift Detection
- Conformal Prediction
- X-fold cross validation
- Continuous Evaluation
- What If simulation
- Sensitivity Analysis
- Signal to noise ratio
- Fast Fourier Transform
- Discrete Wavelet Transform