---
title: "Slides: Respect The Process - Andrew Dumit, Watershed Technology Inc."
category: "slides"
video_id: "CLttOU7n6sI"
sourceLabels: ["Public YouTube video frames", "Public YouTube metadata"]
---

# Slides: Respect The Process - Andrew Dumit, Watershed Technology Inc.

## Source Video
[Respect The Process - Andrew Dumit, Watershed Technology Inc.](https://www.youtube.com/watch?v=CLttOU7n6sI)

## 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
![[assets/slides/CLttOU7n6sI/slide-001.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-001.html)
- AI slide classifier: `title_card` confidence `0.99`
- Text source: agent_vision.

Slide text:

> Respect the Process
> Trustworthy coding agents in a domain full of expert judgement calls

![[assets/slides/CLttOU7n6sI/slide-002.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-002.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: agent_vision.

Slide text:

> Sustainability is filled with judgement calls
> Questions like:
> - What are the emissions attributable to 1 bottle of wine?
> - Which method should be used to allocate the emissions to various co-products?

![[assets/slides/CLttOU7n6sI/slide-003.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-003.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: none.
- OCR decision: ready — Small multi-region text with citation and metric is better suited for OCR than manual transcription in triage.
![[assets/slides/CLttOU7n6sI/slide-004.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-004.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: none.
- OCR decision: ready — Diagram slide with many small labels and node cards is OCR-suitable.
![[assets/slides/CLttOU7n6sI/slide-005.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-005.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: none.
- OCR decision: ready — Diagram slide with repeated node cards and small labels is OCR-suitable.
![[assets/slides/CLttOU7n6sI/slide-006.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-006.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: none.
- OCR decision: ready — Dense repeated diagram nodes and small labels are OCR-suitable.
![[assets/slides/CLttOU7n6sI/slide-007.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-007.html)
- AI slide classifier: `content_slide` confidence `0.97`
- Text source: agent_vision.

Slide text:

> So, we naturally swapped in a coding agent
> Gives the agent the ability to delight
> The agent can find clever ways to solve underspecified problems, create fancy visualizations, and answer related questions

![[assets/slides/CLttOU7n6sI/slide-008.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-008.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: none.
- OCR decision: ready — Three-column text slide is OCR-suitable and should be handled by OCR rather than manual transcription in triage.
![[assets/slides/CLttOU7n6sI/slide-009.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-009.html)
- AI slide classifier: `content_slide` confidence `0.99`
- Text source: agent_vision.

Slide text:

> But, unconstrained code is scary
> The agent will find “creative” ways to solve every problem
> The agent can gaslight users saying it made edits
> Manual review of code is not in our users’ wheelhouse

![[assets/slides/CLttOU7n6sI/slide-010.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-010.html)
- AI slide classifier: `content_slide` confidence `0.97`
- Text source: agent_vision.
- OCR decision: ready — Dense multi-column slide with chart and several small quotes.

Slide text:

> The process ≠ answer story is not new

![[assets/slides/CLttOU7n6sI/slide-011.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-011.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: agent_vision.
- OCR decision: ready — Code-heavy slide with bullets and a large code sample.

Slide text:

> Our SDK as the only door

![[assets/slides/CLttOU7n6sI/slide-012.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-012.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: agent_vision.
- OCR decision: ready — Text plus structured flow diagram and step boxes are OCR-suitable.

Slide text:

> Deterministic execution to guarantee process

![[assets/slides/CLttOU7n6sI/slide-013.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-013.html)
- AI slide classifier: `content_slide` confidence `0.96`
- Text source: agent_vision.
- OCR decision: ready — Mostly the same structured slide with an overlaid report and small labels.

Slide text:

> Deterministic execution to guarantee process

![[assets/slides/CLttOU7n6sI/slide-014.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-014.html)
- AI slide classifier: `content_slide` confidence `0.95`
- Text source: agent_vision.
- OCR decision: ready — Structured process slide with chart labels and small diagram text.

Slide text:

> Deterministic execution to guarantee process

![[assets/slides/CLttOU7n6sI/slide-015.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-015.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: agent_vision.

Slide text:

> Hillclimbing within this harness
> All the typical approaches were still effective
> Prompt improvements
> Few-shot examples
> Making tools better fit-for-use
> Breaking the task down
> Teaching the agent some expert judgement
> 43% -> 92%
> Accuracy improvement on internal evals

![[assets/slides/CLttOU7n6sI/slide-016.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/slides/CLttOU7n6sI/slide-016.html)
- AI slide classifier: `content_slide` confidence `0.97`
- Text source: agent_vision.

Slide text:

> In a domain full of judgement calls, respect the process
> User request
> agent writes free code
> typed SDK + lint
> deterministic execution
> commit as typed objects
> valid · traceable · replayable
> reject → retry


Classification audit: `raw/sources/slide-ai-classification/slides/CLttOU7n6sI/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.