Hands-on AI training for engineering teams

Your engineers already have the tools. This is the day the team agrees on how to use them. By the afternoon, everyone is writing the team's conventions doc together.

• Prompting that holds up in a large, messy codebase
• Rules, skills, and config for the assistant your team uses
• Refactoring and debugging legacy code with an agent
• Reviewing AI-written code: what to trust and what to read line by line
• Running several agents at once without stepping on each other

We spend the day building. Your team picks a real problem, builds an agent to handle it, and tests it properly before it goes anywhere near production. You leave with a working agent and a plan for who maintains it.

• How to tell a useful agent from a prompt with extra steps
• Connecting agents to internal tools, APIs, and data with MCP
• Putting agents where your team already works, like Slack
• Testing and evals for agent behavior
• Cost, latency, and when to pull the plug on an agent that isn't earning its keep

Once everyone is comfortable working with an agent, the next step is work that runs without someone at the keyboard. We design that pipeline with your team: agents pick up tickets, write code in CI, and open pull requests, and people step in at the review points you choose. By the end of the day, you have a design for your pipeline and one workflow running end to end.

• Turning your team's conventions into skills, hooks, and checks agents follow
• Running agents in CI and as background jobs
• Permissions, secrets, and what an agent should never be allowed to touch
• Measuring what the factory ships and what it costs

For product, marketing, and ops folks who work next to engineers and want in on the AI work. Most people leave with a workflow they use the next morning.

• What a model is doing when it answers you
• Using AI for writing, research, and analysis
• How to tell when an AI feature is broken, and what to report
• Where AI creates risk and how to talk about it with your team