Building Intelligent Autonomous Agents

Follow this structured process to build intelligent autonomous agents that outperform existing decision-making methods.

1. Create your agent shell The agent shell is the basic framework of your agent. You will start by creating your shell and then add additional building blocks to it through the process.
2. Define sensors for your agent Sensors provide feedback that your agent will use to learn its task.
3. Add perception to your agent The perception layer of the agent synthesizes additional information that your agent will use to learn.
4. Constructing scenarios for your agent Scenarios are situations in which your agent will practice completing the task.
5. Adding skills to your agent Skills are the key building blocks of intelligent autonomous agents. You will build your skills layer by breaking down the process into skills, orchestrating the skills together, and either programming each skill with math, rules, or algorithms, or teaching it with deep reinforcement learning.
6. Train your agent. The Composabl Python SDK manages the training of your agent.
7. Deploy your agent Once your agent has finished training, you can export it to deploy in your system.