Log in or create a free Rosenverse account to watch this video.
Log in Create free account100s of community videos are available to free members. Conference talks are generally available to Gold members.
Hands-on AI #1: Let’s write your first AI eval
This video is featured in the Evals + Claude playlist.
Summary
If you’re a product manager, UX researcher, or any kind of designer involved in creating an AI product or feature, you need to understand evals. And a great way to learn is with a hands-on example. In this talk, Peter Van Dijck of the helpful intelligence company will walk you through writing your first eval. You will learn the basic concepts and the tools, and write an eval together. This talk is hands on; you can follow along, and there will be plenty of time for questions. You will go away with an understanding of the basic building blocks of AI evals, and with the confidence that you know how to write one. And more importantly, you’ll build some intuition, some product sense, around how the best AI products today are built, and how that can help you use them more effectively yourself.
Key Insights
-
•
Evals consist of a task, a golden dataset with known correct outputs, and an evaluator that measures correctness.
-
•
Manual AI prompt testing is slow and inconsistent; automated evals accelerate and scale evaluation.
-
•
UX and product teams can and should learn evals as a practical, non-technical skill.
-
•
Creating your own golden dataset is essential and cannot be outsourced or fully automated.
-
•
Models are fixed once trained; improvements happen by refining prompts and context design, not retraining the model.
-
•
Evaluations measure task performance, not the underlying model itself, allowing comparison across models.
-
•
Outputting a confidence score from models is unreliable due to lack of internal memory and inconsistent scale interpretation.
-
•
Biases are baked into models during training via evals used in post-training refinement.
-
•
LLMs can be used to judge other LLM outputs to evaluate tasks with non-binary answers.
-
•
Effective eval work requires collaboration across data analysts, engineers, subject matter experts, and UX/product teams.
Notable Quotes
"Evals are like a way to define what good looks like."
"The model was baked and once it’s baked, it does not learn again until they bake a new one."
"You need to be looking at the data. Nobody wants to, but that’s core work."
"Without a golden dataset, you have to build the golden dataset yourself."
"We’re not teaching the model anything; we’re improving our prompts and context."
"Confidence scores from the model are not a good idea because the model has no memory."
"Biases are baked in through the evals used during model training and post-training."
"LLMs judging other LLMs might sound crazy, but if you do it right, it works."
"Evals are a product and UX skill; learning them lets you make these systems do what you want."
"There is a large and growing capability overhang in these models we haven’t discovered yet."
Or choose a question:
More Videos
"Empathy walls emerged at DTA because teams needed a visceral way to bring stories back from the field."
Leisa ReicheltOpening Keynote: Operating in Context
November 7, 2018
"Support structures for switching from design to academia are sparse; you mostly have to figure it out yourself."
Cennydd BowlesExit Interview #2: Rediscovering the ethical heart of design
November 6, 2025
"Embedded squads build domain expertise but risk losing knowledge if someone leaves, while centralized teams help reduce duplicated effort."
Andrew MichaelBuilding a Product Insights Team
March 10, 2022
"Having a back office support in place made it much easier to convince new researchers to join the team."
Veevi RosensteinBuilding for Scale: Creating the Zendesk UX Research Practice
January 8, 2024
"Age carries with it obvious visible differences and lived experience, while race often involves less visible but equally impactful structures."
Megan CamposWhat Did I Miss? The Hidden Costs of Deprioritizing Diversity in User Research
March 12, 2021
"The app reads the question and prepares tutoring material, then helps solve it step by step or through chat if further questions remain."
Aras BilgenWho does the math: A designer’s journey in building an AI-based tutoring app
June 10, 2025
"If you do enough research, you might find 20% of your users are edge cases, so maybe you’re not building for the right people."
Ron BronsonDesign, Consequences & Everyday Life
November 18, 2022
"The chaos of undocumented processes makes it hard to build useful AI workflows; mapping and blueprinting help."
Tina WeisserWhen AI Agents Meet Reality. Service Design Lessons from a Pilot
February 26, 2026
"As the product journey progressed, research cadence got faster with prototypes and quicker feedback loops."
Dr Chloe SharpUsing Evidence and Collaboration for Setting and Defending Priorities
November 29, 2023
