---
schema: 1
type: video
title: >-
  How Evals and Prompts Shape Agent Behavior — Preetika Bhateja & Daniel Bump,
  YouTube Ads
description: >-
  Getting an AI agent to behave the way you want isn't just about writing better
  prompts. In real systems, behavior emerges from a loop: prompts, evals,
  iteration, and feedback. Small changes in any part of that loop can completely
  change outcomes. The Google team shares lessons from building a seed-asset
  agent that turns messy advertising creatives — low-quality images, cluttered
  visuals, and heavy text overlays — into clean, reusable assets for downstream
  generative AI tools. They explain why prompting alone did not produce stable
  behavior, how evals became feedback signals rather than scorecards, how agent
  trace logs exposed why failures happened, and how they iterated without
  breaking problems they had already fixed. Speakers: Chris Souza — Google Chris
  works on the Google team behind this seed-asset agent and its evaluation
  workflow. Preetika Bhateja — Product Manager, Google/YouTube Preetika works on
  ads, evaluations, agents, and LLM-as-judge systems. Daniel Bump — Engineer,
  Google Daniel focuses on image and video generation and computer vision.
  X/Twitter: https://x.com/DanielJBump LinkedIn:
  https://www.linkedin.com/in/danielbump
canonical: >-
  https://liquidgold.cc/channel/aidotengineer/explore/video/how-evals-and-prompts-shape-agent-behavior-preetika-e018daf1bd7b
markdown: >-
  https://liquidgold.cc/channel/aidotengineer/video/how-evals-and-prompts-shape-agent-behavior-preetika-e018daf1bd7b.md
llms: null
parent: 'https://liquidgold.cc/channel/aidotengineer.md'
channel: aidotengineer
updated: '2026-10-24T06:30:21.765Z'
language: null
children: []
videoId: xyL2Ltkh-SA
durationSeconds: 1169
publishedAt: '2026-07-24T21:00:00.000Z'
searchChunks:
  - ordinal: 0
    text: null
    startSeconds: 23
    endSeconds: 162
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=23'
    heldReason: rights
  - ordinal: 1
    text: null
    startSeconds: 23
    endSeconds: 162
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=23'
    heldReason: rights
  - ordinal: 2
    text: null
    startSeconds: 23
    endSeconds: 162
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=23'
    heldReason: rights
  - ordinal: 3
    text: null
    startSeconds: 175
    endSeconds: 277
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=175'
    heldReason: rights
  - ordinal: 4
    text: null
    startSeconds: 175
    endSeconds: 277
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=175'
    heldReason: rights
  - ordinal: 5
    text: null
    startSeconds: 288
    endSeconds: 420
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=288'
    heldReason: rights
  - ordinal: 6
    text: null
    startSeconds: 288
    endSeconds: 420
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=288'
    heldReason: rights
  - ordinal: 7
    text: null
    startSeconds: 430
    endSeconds: 510
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=430'
    heldReason: rights
  - ordinal: 8
    text: null
    startSeconds: 523
    endSeconds: 615
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=523'
    heldReason: rights
  - ordinal: 9
    text: null
    startSeconds: 523
    endSeconds: 615
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=523'
    heldReason: rights
  - ordinal: 10
    text: null
    startSeconds: 615
    endSeconds: 626
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=615'
    heldReason: rights
  - ordinal: 11
    text: null
    startSeconds: 626
    endSeconds: 727
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=626'
    heldReason: rights
  - ordinal: 12
    text: null
    startSeconds: 738
    endSeconds: 823
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=738'
    heldReason: rights
  - ordinal: 13
    text: null
    startSeconds: 834
    endSeconds: 938
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=834'
    heldReason: rights
  - ordinal: 14
    text: null
    startSeconds: 834
    endSeconds: 938
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=834'
    heldReason: rights
  - ordinal: 15
    text: null
    startSeconds: 1050
    endSeconds: 1169
    watch: 'https://liquidgold.cc/watch/xyL2Ltkh-SA?t=1050'
    heldReason: rights
---


# How Evals and Prompts Shape Agent Behavior — Preetika Bhateja & Daniel Bump, YouTube Ads

Getting an AI agent to behave the way you want isn't just about writing better prompts. In real systems, behavior emerges from a loop: prompts, evals, iteration, and feedback. Small changes in any part of that loop can completely change outcomes. The Google team shares lessons from building a seed-asset agent that turns messy advertising creatives — low-quality images, cluttered visuals, and heavy text overlays — into clean, reusable assets for downstream generative AI tools. They explain why prompting alone did not produce stable behavior, how evals became feedback signals rather than scorecards, how agent trace logs exposed why failures happened, and how they iterated without breaking problems they had already fixed. Speakers: Chris Souza — Google Chris works on the Google team behind this seed-asset agent and its evaluation workflow. Preetika Bhateja — Product Manager, Google/YouTube Preetika works on ads, evaluations, agents, and LLM-as-judge systems. Daniel Bump — Engineer, Google Daniel focuses on image and video generation and computer vision. X/Twitter: https://x.com/DanielJBump LinkedIn: https://www.linkedin.com/in/danielbump

Canonical: https://liquidgold.cc/channel/aidotengineer/explore/video/how-evals-and-prompts-shape-agent-behavior-preetika-e018daf1bd7b · Channel: [AI Engineer](https://liquidgold.cc/channel/aidotengineer.md)

> Search chunk text is held until public-source provenance is verified.

## Document chunks

> Missing: no stored document chunks are available.