# Presentation review and handoff

Pretectal workflow template · October 2, 2026
Guide: https://pretectal.com/workflows/review-ai-assisted-presentations
Editable worksheet: https://pretectal.com/tools/presentation-review

Start by freezing a copy of the intended presentation or export. Record its exact version and owner before opening the review worksheet. Use NOT REVIEWED for checks that have not occurred.

## 1. Set the review scope

- Presentation title, version, and date:
- Audience and event:
- Responsible owner:
- Activity type and applicable provider / organizer requirements:
- Scientific review owner:
- Visual / illustration review owner:
- Delivery test owner:
- Frozen file or hosted build to review:

## 2. Inspect claims and visuals

### Issues log

For each issue, record scene / slide, exact concern, supporting source, proposed correction, responsible person, resolution, and version retested. Do not close an issue merely because a model agreed with the draft.

| Issue ID | Scene / asset | Exact concern and source | Correction required | Owner | Open / resolved / needs decision | Resolution evidence and retested version |
| --- | --- | --- | --- | --- | --- | --- |
| | | | | | | |

Keep a material unsupported claim or visual open until corrected, removed, or otherwise resolved by the responsible reviewer. A disclosure is not a resolution.

## 3. Test the delivery

- Exported file or hosted version actually tested:
- Device and delivery setting:
- All response branches and expected feedback compared:
- Keyboard, touch, screen-reader, and static paths checked:
- Broken links, unreadable labels, or fallback differences:
- Corrections and version retested:

### Review record

- [ ] Objective and audience are explicit.
- [ ] Material claims, values, and limitations checked against original sources.
- [ ] Visual relationships and all animation states checked.
- [ ] Asset origins, permissions, and attributions recorded.
- [ ] Each answer, branch, and feedback state tested.
- [ ] Keyboard, readability, accessible alternatives, and static fallback checked.
- [ ] Actual AI contribution documented accurately.
- [ ] Final owner, remaining issues, and correction route recorded.

## 4. Describe the AI contribution

Complete only with facts about the work performed:

We used [tool and version if known] on [date] to [specific task]. The material supplied was [description]. [Person / role] checked [specific output] against [sources or criteria] and made [corrections]. Remaining limitations are [limitations].

Do not invent a reviewer or imply completed clinical review when it has not occurred. Follow the responsible provider's disclosure requirements.

## 5. Close and preserve

### Package manifest

- Editable presentation:
- Static fallback:
- Evidence map:
- Sources:
- Asset register:
- Final review log:
- AI disclosure:
- Change log:
- Correction contact:

### Release decision

- Decision owner:
- Decision and date:
- Exact approved version, if approval has occurred:
- Remaining restrictions or unresolved issues:
- Review triggers:

Completion check: the delivered file matches the tested version and material issues have recorded resolutions. If a required review is missing or the content changes after review, keep release pending and send the affected part back to its owner. Preserve remaining limits with the handoff.

This handoff is not CE accreditation, certification, or an independent assessment of scientific accuracy.