Reconstructed Slides: CIAM for AI: Authn/Authz for Agents — Michael Grinich, CEO of WorkOS

Source Video

CIAM for AI: Authn/Authz for Agents — Michael Grinich, CEO of WorkOS

Method

This deck is reconstructed from the existing video frame captures by detecting likely slide regions with OpenCV, cropping/upscaling those regions, deduplicating similar crops, and OCRing the cropped slide images locally. It is a cleaner companion to the full-stage frame deck.

Reconstructed Slides

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

Slide text:

Good morning!

What should we

build today?

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

Slide text:

AI Agent

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

Slide text:

AI Agent

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 0.92
Text source: agent_vision.
OCR decision: ready — Dense small compliance list and labels likely better handled by OCR.

Slide text:

Compliance

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

Slide text:

Persona Shadowing

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

Slide text:

Delegation Chains

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

Slide text:

Capability-Based Tokens

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

Slide text:

Persona Shadowing

Delegation Chains

Capability Tokens

Escalation to Human

Recreated text/layout view: open HTML recreation
AI slide classifier: content_slide confidence 0.95
Text source: advanced OCR rapidocr-live/bright-screen/opencv-adaptive reconciled by agent.
OCR decision: ready — Dense acronym wall is OCR-suitable and likely cheaper than visual transcription.
Slide text: not surfaced (decorative by AI classifier).

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

Slide text:

GNAP

Grant Negotiation and Authorization Protocol

(RFC 9635)

Recreated text/layout view: open HTML recreation
AI slide classifier: diagram confidence 0.92
Text source: none.
Slide text: not surfaced (none by AI classifier).

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

Slide text:

AuthKit

FGA

Radar

Recreated text/layout view: open HTML recreation
AI slide classifier: other confidence 0.83
Text source: none.
Slide text: not surfaced (none by AI classifier).

Recreated text/layout view: open HTML recreation
AI slide classifier: diagram confidence 0.9
Text source: none.
Slide text: not surfaced (none by AI classifier).

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

Slide text:

Find me on X to discuss

identity for AI agents

@grinich

Hidden Non-Slide Evidence

`slide-001.jpg` — sponsor_logo confidence 0.99; sponsor logo wall
`slide-002.jpg` — title_card confidence 0.96; title card / intro slide
`slide-019.jpg` — speaker_stage confidence 0.98; Speaker on stage with podium; only a cropped, unreadable presentation screen edge is visible.
`slide-020.jpg` — speaker_stage confidence 0.98; Speaker-at-podium shot; the visible slide is only a narrow cropped edge and not readable.

Classification audit: raw/sources/slide-ai-classification/reconstructed/D4Dswf-__RM/audit.json