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.
Building impactful AI products for design and product leaders, Part 2: Evals are your moat
This video is featured in the AI and UX playlist.
Summary
Simply Put’s Peter van Dijck will demystify evals, and share a simple framework for planning for and building useful evals, from qualitative user research to automated evals using LLMs as a judge.
Key Insights
-
•
AI product development involves three layers: model capabilities, context management, and user experience, with evals central to experience quality assurance.
-
•
Automated evals help scale testing of AI with inherently open-ended inputs and outputs, enabling faster iteration cycles with confidence.
-
•
LLMs can serve as judges (evaluators) of other LLM outputs, which works because classification is cognitively easier than generation.
-
•
Defining what 'good' means for an AI system is a detailed, evolving process informed by research, domain expertise, and observed risks.
-
•
A three-option evaluation (e.g., yes/no/maybe) works better than fine-grained scales for consistent automated scoring by LLMs.
-
•
Synthetic data, generated by LLMs based on manually created examples, efficiently expands dataset breadth and usefulness.
-
•
Domain experts are essential for tagging data and establishing quality criteria, especially for high-stakes areas like healthcare or legal.
-
•
Building effective evals requires substantial effort—expect 20-40% of project resources devoted to this work.
-
•
Cultural differences impact subjective evals like politeness, requiring localization and careful domain definition.
-
•
AI product quality management is a strategic ongoing commitment, extending beyond initial development into production monitoring and iteration.
Notable Quotes
"AI products almost always have both open-ended inputs and outputs, which makes testing really hard."
"You have to build a detailed definition of what is good for my system to do meaningful automated evals."
"It’s much easier to classify an answer than to generate an answer, and that’s why LLM as a judge works."
"You don’t want to give too many options like rating from one to ten because consistency gets lost between different LLM calls."
"Synthetic data is useful because it’s easier to generate more examples of something you already have than to create entirely new data."
"If you launch in the US and politeness is an issue, first try to fix it with prompts; only if that fails should you build an eval."
"Evals are really your intellectual property—they define what good looks like in your domain."
"Domain experts are crucial for tagging data because users might say ‘that’s great,’ but experts can tell it’s totally wrong."
"You should plan 20 to 40 percent of your project budget on evals—it’s a lot more work than most people expect."
"This is where UX and product strategy bring huge value—defining what good means rather than leaving it to engineers alone."
Or choose a question:
More Videos
"We encourage one-to-one video calls to build rapport and feel comfortable reaching out for advice."
Marc Fonteijn Ru ButlerIncrease your confidence, influence, and impact (through a Professional Community)
December 3, 2024
"Design’s value is in asking meaningful questions and steering towards the right outcome, not getting bogged down in technical parameters."
Dem Gerolemou Alexis Oh Neef RehmanClimate technology fundamentals
May 15, 2024
"Our benchmarking survey includes free exploration of the entire website or app, not just the homepage."
Daniela Magaña Flores Ariane Rahn Jeff Ephraim BanderAhead of Competition: Learn What UX Benchmarking Can Do for Your Business Today
March 10, 2022
"Design systems are not just components but guidelines, rules, and principles to create consistent experiences."
Abbey Smalley Sylas SouzaScaling UX Past the Size of Your Team
January 8, 2024
"Anything below zero on the design staffing score indicates we’re staffing lower priority projects over higher ones."
Patrick CommarfordDesign Staffing for Impact
January 8, 2024
"Data sovereignty is about who owns your data, which country’s laws apply, and how vulnerable groups are fairly protected."
Xenia Adjoubei Sean BruceEmpowering Communities Through the Researcher in Residence Program
March 29, 2023
"Dark patterns are UI designs that trick or manipulate users into doing something they wouldn't otherwise do or that go against their or society’s best interests."
Craig VillamorDesign Systems for Ethical Design
January 26, 2023
"User research is really only part of the information schema that we’re talking about."
Jon Fukuda Ellie KryslDesign Planning and Management Support
October 3, 2023
"More frequent, open communication and direct conversations helped us avoid assumptions and deliver something usable quickly."
Iain McMaster IHan ChengDesign and Product: from Frenemy to Harmony
November 29, 2023
