SparkCognition Darwin™

SparkCognition Darwin™

SparkCognition's Darwin™ is an automated machine learning (AutoML) product that empowers you with the right set of tools to quickly prototype use cases and achieve results faster than traditional data science methods. Darwin accelerates data science at scale, enabling you to assess the quality of your dataset and advising you on how to fix problems to make it suitable for the model-building process. Darwin then automates time-consuming tasks that range from model creation and optimization to model deployment and continuous maintenance. Darwin uses a patented neuroevolutionary approach to custom-build models that perfectly fit your data, resulting in superior levels of accuracy and performance. Through this blend of evolutionary algorithms and deep learning methods, Darwin accelerates the iterative improvement of models while allowing different degrees of control for customization.

*Please note that member solutions are often customizable to meet the needs of individual enterprise end users.



  • Expedite Data Preparation Tasks
    Darwin proactively uncovers problems within your dataset and interactively provides advice on how to address them, ensuring your data is ready for the automated model-building process.
  • Create and Manage Thousands of Models
    Darwin enables data science and analytics teams to build, deploy, and maintain all of their models at scale from a single productive environment that fully integrates with your existing toolchain through SDK and API options.
  • Control the Whole Process from Data to Model
    Darwin allows complete governance over methods used to prepare your data and selection of optimization functions, helping you influence the results you want while preserving the accuracy and performance of models.
  • Unparalleled Accuracy Through Deep Learning
    Darwin's neuroevolutionary process specializes in the search and auto-tuning of neural architectures based on the intricacies of your data, defining and exposing every aspect of the network and resulting in more accurate, sophisticated models.
  • Handle Complex Temporal Relationships
    Darwin uses long short-term memory (LSTM) and temporal convolutional network (TCN) architectures to capture complex relationships over time and exploit them to make more accurate predictions.
  • True Generalization to Address the Unknown
    Darwin's neuroevolutionary process is based on a fitness function. It quickly adapts to changes in upcoming data to achieve maximum accuracy under dynamic circumstances, evolving to maintain complexity, efficacy, and efficiency.


WorldwideIntel® Xeon ScalableCSP - Google CloudOn-premise (Private Cloud, Other)Torch/PyTorchCross-IndustryEnergy and UtilitiesFinance and InsuranceManufacturingModels can be trained - data input only requiredModels can be trained - online learningModels can be trained - requires labeled dataModels cant be re-trained - Inference onlyAI Software/SaaSLinux Intel® Math Kernel Library for Deep Neural Networks (Intel MKL-DNN)LSTMProprietaryData AnalyticsData Preparations and ManagementPredictive maintenance and analyticsDeep Learning