Top 10: AutoML Tools

Automated machine learning (AutoML) is the process of automating manual tasks that must be completed as data scientists build and train machine learning (ML) models. 

By using AutoML tools, developers and data scientists can automate the time consuming tasks in ML model development. 

Technology company IBM says that AutoML democratises ML by making it accessible to anyone who is interested in exploring its potential. 

AI Magazine has ranked the Top 10 AutoML platforms on offer today across large technology companies and individual platforms, analysis scale, innovative features and usefulness. 

10. MLJAR AutoML

CEO: Piotr Płoński
Company: MLJAR
Headquarters: Łapy, Poland

MLJAR aims to make ML accessible to non-experts while streaming workflows for experienced practitioners and improving the efficiency and performance of ML models.

The company says its platform, MLJAR AutoML, provides advanced automated ML capabilities that enable effective model creation and data analysis process optimisation.

MLJAR's AutoML has been downloaded over one million times and has unique features such as automated documentation and fairness metrics.

9. Watson Studio AutoAI

CEO: Arvind Krishna
Company: IBM
Headquarters: New York, US​​​​​​​

IBM's AutoAI is a variation of AutoML that extends the automation of model building to the entire AI lifecycle. Part of IBM's Watson Studio, it applies intelligent automation to the steps of building predictive ML models.

AutoAI's features include data pre-processing, automated model selection, feature engineering, hyperparameter optimisation, model monitoring integration and model validation support. 

Using AutoAI, IBM says scientists, developers and ML engineers can automatically build ML and AI models without deep data science expertise.

8. KNIME Analytics AutoML Component

CEO: Trevor Kaufman
Company: KNIME
Headquarters: Zurich, Switzerland

KNIME is a single platform for end-to-end data science. Beyond building workflows, commercial teams leverage KNIME to ensure sensitive data is kept safe and that AI models are validated and monitored.

The KNIME Analytics platform features a component that automatically trains supervised ML models for both binary and multiclass classification. 

The component is able to automate the whole ML cycle by performing some data preparation as well as parameter optimisation with cross validation.

It also performs scoring, evaluation and selection.

7. RapidMiner

CEO: Roland Busch
Company: Siemens
Headquarters: Munich, Germany

Owned by parent company Siemens, Altair's RapidMiner is an integrated pillar within Siemens Xcelerator portfolio. 

Rapidminer is a powerful enterprise AI and analytics solution that connects siloed data, uncovers hidden insights and provides advanced analytics and AI automation.

Siemens says that by using RapidMiner, businesses can build predictive models in a few clicks while leveraging automated functionality, including AutoML.

RapidMiner says its AutoML tool is good at finding non-obvious relations and relevant data.

6. Dataiku DSS

CEO: Florian Douetteau
Company: Dataiku
Headquarters: New York, US​​​​​​​

Founded in France and now headquartered in the US, Dataiku says data scientists and analysts alike can build and compare production-ready models quickly and with “white-box explainability.”

Dataiku's data science studio (DSS) contains a powerful automated ML engine that allows users and businesses to get highly optimised models with minimal intervention.

Dataiku says its platform combines the simplicity of autoML for fast prototyping with more advanced visual ML capabilities for creating sophisticated models.

5. H2O AutoML

CEO: Sri Ambati
Company: H2O.ai 
Headquarters: California, US

H2O.ai provides an end-to-end gen AI platform. It works with partners across the business sector and academia.

Its AutoML can be used for automating the ML workflow, which includes automatic training and tuning of many models within a user-specified time-limit.

H2O offers a number of model explainability methods that apply to AutoML objects (groups of models), as well as individual models. 

Explanations can be generated automatically with a single function call, providing a simple interface to exploring and explaining the AutoML models.

4. SageMaker Autopilot

CEO: Matt Garman
Company: AWS
​​​​​​​Headquarters: Washington State, US

Part of the Amazon SageMaker portfolio of AI tools and capabilities, Autopilot automatically trains and tunes ML models for classification or regression, based on a business' data - while allowing full control and visibility.

AWS says that its Autopilot now can provide integrated data preparation, multi-modality support, built-in visualisations, “what-if analysis” and automation support for predictions.

Using the service a user can deploy the model to production with just one click or iterate with the recommended models in AWS SageMaker Canvas.

3. Azure Machine Learning AutoML

​​​​​​​CEO: Satya Nadella
Company: Microsoft
Headquarters: Washington State, US

US technology giant Microsoft offers AutoML services as part of Microsoft Azure, its cloud computing platform. 

Microsoft says that users or businesses can create ML models quickly with no-code UI and SDKs. 

During training, Azure Machine Learning creates many pipelines in parallel that try different algorithms and parameters. 

The service iterates through ML algorithms paired with feature selections. 

Microsoft's Azure service also has automated featurisation techniques for data exploration and pre-processing.

2. Google Cloud Gemini Enterprise Agent Platform

CEO: Sundar Pichai
Company: Google
Headquarters: California, US

Formerly called Vertex AI, Google Cloud's Gemini Enterprise Agent Platform is a platform for developers to build, scale, govern and optimise agents. 

Gemini Enterprise Agent platform is a massive ecosystem for AutoML and generative AI solutions. 

The platform has a variety of options for agent building, including: Agent Platform to build and scale agents; Agent Studio for large generative AI models; Model Garden for discovering open-source models.

It also features custom training for control over an ML framework. Using the service, Google says its users can build production-ready generative AI agents and applications on a platform that scales. It features more than 200 Google and third-party AI models and tools. 

The platform provides purpose-built MLOps tools for data scientists and ML engineers to automate, standardise and manage ML projects.

The company says that AutoML on Gemini Enterprise Agent Platform provides a way to train high-quality ML models with minimal effort and ML expertise.

1. DataRobot

CEO: Debanjan Saha
Company: DataRobot
​​​​​​​Headquarters: Massachusetts, US

Founded in 2012, DataRobot has pioneered some incredibly important applied technological advancements in AI and ML.

The Boston headquartered company claims it invented both AutoML and Automated Time Series while being a trailblazer in MLOps and Gen AI.

Using DataRobot’s AutoML Platform users can build specialised workflows including data/time partitioning, unsupervised learning with unlabelled or partially labelled data, composable ML, visual AI for image-based datasets as well as location AI for geospatial datasets. 

It has expanded from just offering AutoML. Now, DataRobot is an end-to-end platform that lets businesses build, operate and govern AI agents across any cloud or on-premises environment.

The company claims some of the organisations that use its offerings include FordDirect, the US Army and Razorpay.

DataRobot’s Agent Workforce Platform is fully integrated with the NVIDIA Enterprise AI Factory and is validated to deliver a complete, production-grade agentic AI stack.