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.
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
"The more experienced you are, the more satisfied you tend to be with your salary; it can almost be double compared to beginners."
Marc FonteijnFirst Insights from the 2025 Service Design Salary(+) Report
December 4, 2024
"Agents do a lot of jobs simultaneously with completely different mental contexts."
Sean Fitzell Sarah Han Kayla FarrellCraft of User Research: Building Out Jobs to be Done Maps
March 12, 2021
"All commercially available facial recognition software perform worse on darker females."
Dr. Jamika D. BurgeBroad Strokes: Connecting Design, Research, and AI to the World Around Us
June 7, 2023
"Everyone contributes to conflict; it’s not about right or wrong, but what each person is trying to protect."
Laura WeissTurn Down the Heat: 3 Ways to Handle Conflict in the Moment
November 20, 2024
"The pandemic flipped the equation, and IBM became the place that design talent left."
Dante GuintuHow to Crush the Talent Crunch
September 8, 2022
"I want people to question why I’m not in the room if I’m not there."
Renee ReidBecoming a ResearchH.E.R (Highly Enterprise Ready)
June 3, 2019
"My younger self had a lot of impostor syndrome, but everyone even with more experience doesn’t totally know what they’re doing."
Sarah Auslander Betsy Ramaccia Gordon RossInsights Panel
November 18, 2022
"Many new adopters use service design artifacts as a tick-box exercise rather than embracing the mindset change required."
Lavrans Løvlie Ben ReasonAsk me anything – Authors of Service Design: From Insight to Implementation
November 19, 2025
"AI development is often uninformed and hurried, resulting in deployments that don’t operate well in the real world."
Joel BranchHumanizing AI: Filling the Gaps with Multi-faceted Research
March 11, 2021
