Skymind Platform | Skymind

Skymind Platform

Enterprise Machine Learning Platform For Teams That Ship


Skymind is an ML platform that guides engineers through the entire workflow of building and deploying machine learning models for enterprise applications on JVM infrastructure. It can be installed on any environment: cloud, on-premises, bare-metal, or hybrid systems. For those that rely on Cisco, Skymind ships with Cisco Hyperflex hardware.

Skymind Machine Learning Platform for Enterprises

Data Adaptation

Extract, transform, and load (ETL) data from storage or streaming sources for use in model training and in production. Native integrations include Spark, Hadoop, Kafka, Cassandra, and Amazon EMR.

Model Development

Use the built-in, Java-based framework to develop, train, and test models in a version-controlled and collaborative environment. Engineers working alongside data scientists can import Python-based notebooks and models such as Keras and TensorFlow for use in JVM environments.

Model development in Skymind

Model Operations

Expose developed models and transformations to enterprise applications via REST API endpoints or the Java client. Monitor and report on performance and resource usage. Retrain, test, and promote models into production.

Model Serving
Embeddable Model Hosting, Management, and Version Control
Multi-Node support
Distribute training and inference across clusters of servers
Scalability & Resiliency
Fault tolerance, load balancing, and leader election
Hardware Acceleration
Managed CUDA for GPU and MKL for CPU

Made for Teams that Ship

Skymind lets engineers work in the language they know best (Java), in existing enterprise environments (JVM), in their interface of choice (CLI or notebook).

          import skil_client
          uploads = client.upload("tensorflow_rnn.pb")
          new_model = DeployModel(name="recommender_rnn", scale=30, file_location=uploads[0].path)
          model = client.deploy_model(deployment_id, new_model)
          ndarray = INDArray(array=base64.b64encode(x_in))
          input = Prediction(id=1234, prediction=ndarray, needsPreProcessing=false)
          result = client.predict(input, "production", "recommender_rnn")

On-Call Machine Learning Experts

Get platform and ML support from the creators of Deeplearning4J—the most popular deep learning library for Java—and one of the top contributors to Keras, one of the most popular deep learning libraries for Python.

Ask an Expert

Schedule a 30-minute demo and Q&A with our enterprise Machine Learning experts.

Talk to a Machine Learning Solutions Expert