Unique features of omega|ml:
Instant deployment: With omega|ml, deployment of ML models, pipelines, datasets, full apps takes seconds instead of hours, weeks or even months. No engineering required, it just works. Cloud based REST and Python APIs are instantly available (instant = as soon as it is saved, there is zero wait time).
Single line of code: A single line of code will deploy scikit-learn, Tensorflow, Keras, PyTorch and any other Python-serializable model. Same for datasets, notebooks and any pip-installable Python package, saving a dataset or a script is a single line of code. Also larger than memory dataframes and other datasets are possible out of the box. It works from laptop to cloud, with any cloud.
Automatic model versioning: Every model is automatically versioned, and accessing any version is as simple as specifying its version tag. The "model as data, not code" means models are stored in the database, not as part of a deployed docker image like everyone else is doing it (treating models as code is fundamentally flawed - after all, models are data, not code).
Extensible:omega|ml is easy to extend thanks to its plugin architecture. Actually most of the standard features are built as plugins.
Also omega|ml is built on well known and wide-spread open source technology, namely Python, PyData, RabbitMQ and MongoDB. This means it is easy to run anywhere, using existing skills and infrastructure.
Open source, Apache 2.0 licensed.
Don't forget to tag @omegaml in your comment, otherwise they may not be notified.