Automated AI
We're building tools to help AI creators reduce the time they spend designing their models. Our goal is to allow non-experts across industries to build their own AI solutions, without writing complex code or performing tedious tuning and optimization.
Our work
How IBM is helping a major retailer stay ahead of the holiday crunch
Case studyKatia Moskvitch- Automated AI
- Machine Learning
Goal-oriented flow assist: supporting low code data flow automation with natural language
Technical noteKartik Talamadupula and Michelle Brachman- AI
- AI for Business Automation
- Automated AI
- Human-Centered AI
Snap ML pushes AutoAI to deliver 4x-faster automated machine learning on IBM Cloud
ReleaseThomas Parnell, Haris Pozidis, Łukasz Ćmielowski, and Daniel Ryszka5 minute read- Automated AI
Simplifying data: IBM’s AutoAI automates time series forecasting
ReleaseSyed Yousaf Shah, Wesley Gifford, and Dhaval Patel6 minute read- AI
- Automated AI
Researchers can speed up their AI model training with Snap ML
ReleaseWill Roberts3 minute read- Automated AI
- Machine Learning
AI for AI set to make it easy to create machine learning algorithms
ResearchKatia Moskvitch8 minute read- AI
- Automated AI
- Machine Learning
Tools + code
Lale: a library for semi-automated data science
A Python library that makes it easy to automatically select algorithms and tune hyperparameters of pipelines that are compatible with scikit-learn, in a type-safe fashion.
View project →Snap ML
Snap ML is a library that helps data scientists in a Python stack accelerate the training and inference of popular ML models.
View project →AutoMLPipeline.jl
A package that makes it trivial to create and evaluate machine learning pipeline architectures, leveraging the built-in macro programming features of Julia.
View project →Lale.jl
A Julia wrapper of Python's Lale library for semi-automated data science.
View project →IBM Federated Learning - Community Edition
A Python framework for federated learning in an enterprise environment.
View project →DOFramework
A testing framework for decision-optimization (DO) model learning algorithms.
View project →
IBM Solution: AutoAI on IBM Watson Studio
Our recent work was developed into AutoAI in IBM Watson Studio. It enables data scientists to quickly build and train high-quality predictive models, and simplifies AI lifecycle management in a code-optional environment.
Publications
- Remo Christen
- Salomé Eriksson
- et al.
- 2023
- ECAI 2023
- Harsha Kokel
- Junkyu Lee
- et al.
- 2023
- IJCAI 2023
- Ibrahim Abdelaziz
- Julian Dolby
- et al.
- 2023
- IJCAI 2023
- 2023
- IJCAI 2023
- Shreyas Basavatia
- Shivam Ratnakar
- et al.
- 2023
- IJCAI 2023
- Cameron Allen
- Timo Gros
- et al.
- 2023
- IJCAI 2023
Tech Preview: IBM Federated Learning
Our research has been developed into a technology preview on the IBM Cloud Pak for Data. Federated Learning provides the tools for training an AI model collaboratively, by using a federated set of secure data sources.