Behavioural Evals AI Workshop

​Ship AI products with confidence.

​Bring one AI-enabled feature and leave with a shared definition of good behaviour and a practical plan for testing it.

​Join Evan Willms for a hands-on, half-day workshop on building the foundations of a Behavior Harness: a repeatable way to check that your AI product behaves as intended as your prompts, models, and product evolve.

​Designed for product owners, engineers, and domain experts building or improving AI-enabled products. Attend individually or with teammates who bring different perspectives on what your product needs to do.

​We’ll work through one continuous exercise, starting with simple behaviour notes and turning them into clear evaluation specifications for your chosen feature.

​You’ll leave with a Behavior Harness Starter Pack, including:

• ​A map of the inputs, context, and system state that influence your feature.

• ​A behavioural map defining what it should do and the boundaries it should respect.

• ​Three to five evaluation specifications describing what to test and what acceptable behaviour looks like.

• ​Prioritised next steps for implementing checks and addressing gaps.

​$400 CAD per person, including the live half-day workshop, self-guided workbook, starter guide, and companion files to help you prepare and continue afterwards.

​Bring your laptop and one feature you’d like to work on.