Scroll to top


Microsoft Bonsai is a low-code AI development platform for intelligent control systems for the speed creation of AI-powered automation

Enable engineers to create AI without requiring data science

Teach AI with instructions and guidance created by engineers

Prepare AI for real environments through simulation

Deploy AI to work independently or in partnership with people

Through a combination of machine teaching, reinforcement learning, and simulation you can build the AI brains that power intelligent control systems.

Microsoft Bonsai is a platform that enables the quick building of intelligent control systems that make recommendations and real-time decisions for equipment or processes within a physical environment.

Machine teaching breaks complex problems into individual skills and gives the brain important information about how to learn faster.

Brains can be deployed in three different ways. In the cloud as a service, on a device integrated directly into an interoperable controller computer, and on the edge using a companion computer that can communicate in real-time with the controller computer

Reinforcement learning takes place in a safe and cost-effective simulated environment removing the risk of damaging a system or taking critical equipment offline. The brain also improves its decision over time to maximise its reward and arrive at a solution that works.

Simulated environments can replicate millions of different real-world scenarios a system might encounter so the brain can learn how to adapt. Brains can also be trained quickly by running these simulations in parallel on Microsoft Azure.

The Microsoft Project Bonsai solution enables engineers to innovate their most dynamic systems and processes with intelligent control systems.

They can apply their expertise to build and manage next-generation control systems in a manner that meets their operational needs. This transformative innovation unlocks new possibilities and drives significant improvements in throughput, efficiency, and quality.

Brains provide decision support in some scenarios and can be given direct decision authority in others. In a decision-making scenario, the solution integrates with existing monitoring software to provide recommendations and predictions to drive consistency among operators and give expert-level insight. With direct decision-making authority, AI brains can develop creative solutions to challenging situations.

Rail Yard that used Deep Reinforcement Learning
Pedestrian simulation and service optimisation
Cat controller with mouse as reinforcement learning
service optimisation for customers beginning
service optimisation for customers middle
service optimisation for customers end

Let’s Collaborate

Ready to
work with us?

We use cookies to give you the best experience.


View Project