---
title: "Designing Evals That Earn User Trust"
category: "talks"
date: "2026-06-29"
time: "1:30pm-1:50pm"
track: "Expo Stage 3 SW"
room: "Expo Stage 3 SW"
speakers: ["Felipe Blanes"]
sourceLabels: ["Official conference schedule", "Public YouTube metadata"]
scheduleTrack: ""
scheduleRoom: "Expo Stage 3 SW"
scheduleLabels: ["Expo Stage 3 SW", "session", "confirmed"]
---
# Designing Evals That Earn User Trust

## Conference Context
- Date/time: 2026-06-29 · 1:30pm-1:50pm
- Track/room: track TBD · Expo Stage 3 SW
- Speaker(s): Felipe Blanes
- Session type/status: session · confirmed

- Track: track TBD
- Room: Expo Stage 3 SW
- Session type: session
- Status: confirmed

## Session Description
Most teams measure their agent against a benchmark, ship it, and hope. But when your agent serves real users, a benchmark won't tell you if it's actually working. This session is about building an eval suite that captures what success looks like in production, runs against real user workflows, and feeds back into product decisions. Here's the flywheel we use in practice: start with what success looks like from the user's perspective, instrument production workflows to capture those signals, diagnose where the agent falls short, and feed those insights into the next thing you build. You'll see how it shaped concrete product bets, turning eval results from a report card into a discovery tool.

## Media Evidence
No related AI Engineer channel video found yet.

## Evidence Graph
This evidence graph is generated from currently linked source material: official schedule text, related video pages, cached transcripts, visible slide text, dense/reconstructed slide pages, and AI slide-classification audits.

### Media Signals
No linked video, transcript, or slide source has been attached yet.

### Agent Reading Notes
Use these signals to refine the synopsis, topic links, people/company context, and method notes. If a source is a related external video rather than an exact official recording, keep it framed as supporting evidence.

## Transcript Status
No official session recording transcript was found by exact title match on the AI Engineer YouTube channel during this run.

## People
- [[felipe-blanes]]

## Notes
- Pending transcript synthesis when an official recording or confirmed matching video is available.

## Synthesis
### Synthesized Breakdown
# Designing Evals That Earn User Trust ## Conference Context - Date/time: 2026-06-29 · 1:30pm-1:50pm - Track/room: track TBD · Expo Stage 3 SW - Speaker(s): Felipe Blanes - Session type/status: session · confirmed - Track: track TBD - Room: Expo Stage 3 SW - Session type: session - Status: confirmed ## Session Description Most teams measure their agent against a benchmark, ship it, and hope. But when your agent serves real users, a benchmark won't tell you if it's actually working. This session is about building an eval suite that captures what success looks like in production, runs against real user workflows, and feeds back into product decisions. Here's the flywheel we use in practice: start with what success looks like from the user's perspective, instrument production workflows to capture those signals, diagnose where the agent falls short, and feed those insights into the next thing you build.

### Speaker And Company Context
- [[felipe-blanes|Felipe Blanes]] — role not listed at [[amazon|Amazon]].

### Topics Covered
- [[agent-security]]
- [[agentic-search]]

### Derived Links And Source Material

### Novel Concepts / Clever Methods
- No highlighted novel concept has been detected yet.

### Evidence Boundary
This synthesis is based on the official schedule and linked source pages. It should be revisited when exact session recordings or transcript-backed secondary sources are available.