Evals for Taste: Hill-Climbing a Slide-Generation Agent
Built rubric-driven replayable eval system from real user projects giving quality/cost/latency/error/token signals in <6 hours per model change. Evolved into dev flywheel powered by real user dissatisfaction signals.
Details
City
London, UK
Date
20 May 2026
Time
1:00PM – 1:45PM
Speaker(s)
Jiri De Jonghe
Anthropic
Agenda
08:00AM – 09:00AM
Check-in and breakfast
9:30AM – 9:45AM
Community general session
·
Boris Cherny
morning sessions
10:00AM – 10:30AM
10:00AM – 10:45AM
10:45AM – 11:15AM
When Code Is a Commodity, What Actually Matters in Product Development?
Builder stage
·
Jacob Langvad Nilsson
11:00AM – 11:45AM
11:30AM – 12:00PM
Building the best agentic analytics harness: Powered by Claude, built with Claude Code
Founder stage
·
Chris Merrick
12:00PM – 12:45PM
12:00PM – 01:20PM
Lunch
afternoon sessions
1:00PM – 1:45PM
1:20PM – 1:50PM
2:00PM – 2:45PM
2:05PM – 2:35PM
2:50PM – 3:20PM
From Idea to Income: How I Built an AI-Powered Murder Mystery Game Business with Claude
Builder stage
·
Michael Whelehan
3:00PM – 3:45PM
3:35PM – 4:05PM
Personality goes a long way: 3 horrible mistakes building a helpful AI notepad
Founder stage
·
Max Tatton-Brown
No items found.
Evening
03:45PM – 06:00PM
Closing reception
Anthropic's developer conference
Join us for a day of hands-on workshops, live demos of new capabilities and conversations with the teams behind Claude. Watch live from anywhere.