Dataiku brings its machine learning platform to the cloud

Get a free Techzine subscription!

Dataiku announced a managed cloud version of its artificial intelligence development platform. The value proposition for this is that it will reduce the amount of work users have to do when building custom machine learning models.

The New-York based company is backed by nearly a quarter of a billion dollars in funding from investors like CapitalG and Google’s parent company, Alphabet. Dataiku achieved its unicorn status in 2019, after the second round of financing and provides a set of tools used by companies to create machine learning models and perform other related tasks like training datasets.

A leap forward

Until now, the software could only be run if customers implemented customized infrastructure environments on-premises or in the public cloud. Dataiku Online is the new managed edition that will make the work easier, by removing the need to maintain infrastructure.

Building an AI model is a two-step process that starts with finding a machine learning algorithm that works for your use case and then refining it by training it using large amounts of data before deployment. The company is now providing users with choices on how to build a model.

What users can do now

Customers can create their own custom algorithm or let the platform choose one using a feature called AutoML. The feature works by examining the type of data the customer wants to process with its AI software and find an algorithm that will efficiently power the process.

The algorithm is selected from a built-in library of AI models that are then refined through training with data provided by the user. There are different offers for companies of all sizes so that even small startups that were founded less than five years ago can access the service.