Markdown source

Slides: The Future Is Domain-Specific Agents - Justin Schroeder, StandardAgents

Source Video

The Future Is Domain-Specific Agents - Justin Schroeder, StandardAgents

Relationship To World's Fair 2026

These slides are extracted from a public AI Engineer YouTube video connected to World's Fair 2026. Speaker-matched clips are supporting context unless later confirmed as exact session recordings; official livestream recordings are day-level/event-level source material.

Related Scheduled Sessions

• No individual scheduled session mapping has been assigned yet; treat this as an event livestream deck.

Extracted Slides

• Recreated text/layout view: open HTML recreation
• AI slide classifier: title_card confidence 0.97
• Text source: agent_vision.

Slide text:

The Future Is Domain-Specific Agents

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.98
• Text source: agent_vision.

Slide text:

We'll harness intelligence with agents.

• Recreated text/layout view: open HTML recreation
• AI slide classifier: title_card confidence 0.99
• Text source: agent_vision.

Slide text:

What is an agent?

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.76
• Text source: agent_vision.
• Slide text: not surfaced (decorative by AI classifier).

• Recreated text/layout view: open HTML recreation
• AI slide classifier: title_card confidence 0.99
• Text source: agent_vision.

Slide text:

But everyone is building agents

• Recreated text/layout view: open HTML recreation
• AI slide classifier: title_card confidence 0.99
• Text source: agent_vision.

Slide text:

Why?

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: agent_vision.

Slide text:

Agents are hard

• Agentic loop orchestration

• Provider abstraction

• Durable execution

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: agent_vision.

Slide text:

It's a mess out there

• Building robust agents is hard.

• There is no defined way to do it.

• Telemetry/observability is hard.

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.98
• Text source: advanced OCR rapidocr-live/bright-screen/opencv-adaptive.
• OCR decision: ready — Dense comparison table with many small labels; OCR is likely more reliable than manual transcription.

Slide text:

Client Resources Prompts Toots Discovery Sampling Roots Elicitation Instructions

Sire x x 0 7 x x 2.?

AgentAl x x? x ×??

AgenticFlow 0 0 x x??

AIQL TUU! x D?

Amazon Q CLI ×? x x??

Amazon Q IDE x × x x x?

Amp 0 D x??

Tester Apify MCP x x x ×?.?

Augment Code x x x x 2?

Avatar Shel! 0 x x x x?

BeeA! × x × x x

Framework

BoltAl x x? x x

Call Chirp Chatbox x x x x x x x x x C

ChatFrame x D x x x

ChatGPT X X X X X

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.95
• Text source: none.
• Slide text: not surfaced (none by AI classifier).

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.96
• Text source: agent_vision.

Slide text:

System Prompt

Model

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.96
• Text source: agent_vision.

Slide text:

Messages

MCP

Skill

Tools

System Prompt

Context

Model

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.95
• Text source: none.
• Slide text: not surfaced (decorative by AI classifier).

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.97
• Text source: advanced OCR rapidocr-live/bright-screen/contrast.
• OCR decision: ready — Dense diagram with small inherited skill/tool labels; OCR is better suited than manual transcription.

Slide text:

iMicr

KAYAK NAVAN

Figma Playwright

Gmail 田 Google Sheets Inheritance

Skill

React Code Fix& Linter

Grill Me

Github

Tools

System Prompt

Model

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.95
• Text source: agent_vision.

Slide text:

Messages

Tools

System Prompt

Figma

Model

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.98
• Text source: advanced OCR rapidocr-live/bright-screen/contrast.
• OCR decision: ready — Multi-node inheritance diagram with repeated small labels and logos; OCR should extract this more reliably.

Slide text:

Messages

Tools

System Prompt

Model

Messages Messages Messages Messages

Tools Tools Tools Tools

System Prompt System Prompt System Prompt System

8Figma Gmail NAVAN Goo

Model Model Model M

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.82
• Text source: none.
• Slide text: not surfaced (none by AI classifier).

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: agent_vision.

Slide text:

Domain Specific Agents...

Far more efficient with tokens

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: agent_vision.

Slide text:

Domain Specific Agents...

Far more efficient with tokens

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: agent_vision.

Slide text:

Domain Specific Agents...

Far more efficient with tokens

Make small language models practical

DeepSeek v4 flash is 137x cheaper than Fable per task

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: advanced OCR rapidocr-live/bright-screen/opencv-adaptive.
• OCR decision: ready — Contains a dense code screenshot; OCR should capture the small embedded text more reliably than manual transcription in this pass.

Slide text:

Domain Specific Agents...

Far more efficient with tokens

Make small language models practical

Can enforce strict limits on capabilities

c> bypass permn ssions on (shift+tab to cycle)

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: agent_vision.

Slide text:

Domain Specific Agents...

Far more efficient with tokens

Make small language models practical

Can enforce strict limits on capabilities

Have excellent scaling characteristics

Hidden Non-Slide Evidence

• `slide-002.jpg` — sponsor_logo confidence 0.99; speaker intro / logo wall, not a presentation content slide
• `slide-003.jpg` — other confidence 0.92; decorative painted background with speaker inset and no readable slide text

Classification audit: raw/sources/slide-ai-classification/slides/spNAUEgq_A8/audit.json

Slide-Derived Subjects To Review

Subject extraction uses video title, related session titles/descriptions, transcript context, and OCR text when available. OCR is best-effort and should be reviewed against the embedded slide images.