Pick a topic to browse its resources, or follow the path for your role. Both paths share the fundamentals and safety & security.
A path is a suggested order through the topics above for your role. Each stop opens that topic in the library below.
Find a useful starting point, build a repeatable workflow, and learn how to evaluate the results.
Suggested reading: AI for Your Small Business: What It Can Do in Your Industry — and How to Start →
Develop software with AI coding agents, then learn to design, evaluate, and operate agent systems.
Suggested reading: AI Agent Engineering: Building Agentic Systems for Enduring Value →
Everything in the collection, newest first. Pick a topic above to narrow it.
32 resources
macOS-specific
Optional project · hands-on work uses code and terminal commands
8-part series · Python · optional, recommended
Practice applying AI to business tasks, checking its output, and deciding where human judgment is needed.
For your business, your team, or the community you serve.
Certification program in development
We're developing Certified Agenteer, Agenteer's upcoming certification program. Explore the learning paths available now, and get updates as we publish the requirements and launch details.
Thanks for subscribing.
Certification news only. Separate from the newsletter.Start by looking across the business, then choose one useful workflow to prove. Compare value, feasibility, and risk before reaching for a tool. The first implementation can be small without being chosen in isolation.
Put AI to work if you run a business or a team and want a workflow that pays off. The Agentic engineering path starts with developing software using coding agents, then covers designing and operating agent systems. The material overlaps, and nothing stops you reading both.
Not for the business path. It assumes you know your own operation and can judge whether an output is good, not that you can code. The engineering path assumes you write software.
You can improve an agent by using results and feedback to refine its instructions, reference material, tools, or workflow. For example, you might clarify a refund rule, then test when a support agent should answer and when it should hand over to a person.
Changing the underlying AI model can also help. Test changes on representative tasks, including work the agent already handles well.
Not yet. Current learning paths and workshops don't award the Certified Agenteer credential. The requirements will be published separately.