Project Bonsai Preview
Create intelligent industrial control systems using simulations
Low-code AI development platform for intelligent control systems
Speed the creation of AI-powered automation to improve production efficiency and reduce downtime—without requiring data science.
Enable engineers to create AI without requiring data science
Teach AI with instructions and guidance created by engineers
Prepare your AI for real environments through simulation
Deploy AI to work independently or in partnership with people
Enable engineers to build AI-powered automation
Start with templates and an intuitive experience that helps you create AI with minimal data science and coding. Choose where to start by optimising a process variable or go big by adding intelligence to an entire process. Safely and quickly validate your AI in a simulated environment before deploying it in the real world.Explore autonomous systems
Improve operator decision making with support from human-trained AI
Infuse the expertise of your engineers have directly into your AI through lessons, goal setting, and rewards. Build once and then reuse your AI in multiple projects to save time and money. Add in your safety policies to help ensure you can meet safety and compliance regulations and keep your staff safe.Learn more about machine teaching
Deploy AI that you have full control over
Easily understand exactly why decisions were made with a black-box-free AI that allows you to better respond to regulators and auditors. You decide how you implement AI - by choosing to use it alongside people for decision support or having it work on its own independently. Best meet your infrastructure needs by choosing to deploy AI on-premises, in the cloud, to the IoT Edge, or in embedded devices.Learn about getting started
Comprehensive security and compliance, built in
- Microsoft invests more than USD 1 billion annually on cybersecurity research and development.
- We employ more than 3,500 security experts completely dedicated to your data security and privacy.
- Azure has more compliance certifications than any other cloud provider. View the comprehensive list.
Project Bonsai resources
Demos and guides
Frequently asked questions about Project Bonsai
Machine teaching is a complementary approach to machine learning. It helps those without AI expertise break a complex problem into simpler tasks and give the AI model important clues about how to find a solution quickly. Machine teaching also makes it easier to understand and audit the autonomous control system’s behaviour after the system has been deployed, which is crucial for safety-critical applications.
Traditional controllers operate on a fixed set of instructions. While these systems effectively perform one task at a time, human operators must manually retune machine settings for different scenarios, conditions, or goals. Additionally, existing technologies are capable of focusing on only one optimisation goal at a time, such as maximising throughput or minimising energy usage.
The platform uses the steps defined in the machine teaching process to inform the deep reinforcement learning models—a machine learning technique in which AI learns by executing decisions and receiving rewards for actions that get it closer to an end goal. Machine teaching accelerates and improves the training process and allows you to reuse the individual steps for other brains.