---
title: "Building an Agentic Video Editor for Mass Consumer"
category: "talks"
date: "2026-07-01"
time: "11:40am-12:00pm"
track: "Generative Media"
room: "Track 1"
speakers: ["Ekaterina Deyneka"]
sourceLabels: ["Official conference schedule", "Public YouTube metadata"]
scheduleTrack: "Generative Media"
scheduleRoom: "Track 1"
scheduleLabels: ["Generative Media", "Track 1", "session", "confirmed"]
---
# Building an Agentic Video Editor for Mass Consumer

## Conference Context
- Date/time: 2026-07-01 · 11:40am-12:00pm
- Track/room: Generative Media · Track 1
- Speaker(s): Ekaterina Deyneka
- Session type/status: session · confirmed

- Track: Generative Media
- Room: Track 1
- Session type: session
- Status: confirmed

## Session Description
Most agentic systems today are built for developers — people comfortable setting up environment, configs, and debugging agent loops. But what happens when your user has never heard the word "agent" and just wants a video ready to post? Reelful is an agentic video editor that lives right in the user's phone. It turns raw photos and videos from your camera roll into polished, short videos. No setup. No sophisticated prompting. No empty timeline. Under the hood, the agent orchestrates multiple models and composes a video together. In this talk, I'll walk through:

- The agentic pipeline architecture: how we chain models across modalities (vision → language → speech → video), handle context passing between steps, and manage state across a multi-minute generation job
- The UX inversion: how we designed the agent to require minimal effort from user — the system infers intent from the media itself, making complex orchestration invisible

This talk is for anyone building agents that need to work for non-technical users, or anyone curious about multimodal agentic pipelines beyond text and code.

## Media Evidence
No related AI Engineer channel video found yet.

- [[youtube-AheG9p_JXVw-transcript]] — full cached transcript markdown for the related YouTube source.

- Source video: `youtube-AheG9p_JXVw`
- Slide deck: [[youtube-AheG9p_JXVw-slides|Slides: AI Engineer World's Fair: Building Reelful - Agentic Video Editor]] — no readable content slides after AI classification.

## Evidence Graph
This evidence graph is generated from currently linked source material: official schedule text, related video pages, cached transcripts, visible slide text, dense/reconstructed slide pages, and AI slide-classification audits.

### Media Signals
- `youtube-AheG9p_JXVw` — 1,340 transcript words
- Transcript signals for `youtube-AheG9p_JXVw`: editing, media, editor, case, real, users, user, everything.
- Evidence links for `youtube-AheG9p_JXVw`: [[youtube-AheG9p_JXVw]], [[youtube-AheG9p_JXVw-transcript]], [[youtube-AheG9p_JXVw-slides]]

### Agent Reading Notes
Use these signals to refine the synopsis, topic links, people/company context, and method notes. If a source is a related external video rather than an exact official recording, keep it framed as supporting evidence.

## Transcript Status
No official session recording transcript was found by exact title match on the AI Engineer YouTube channel during this run.

## People
- [[ekaterina-deyneka]]

## External Secondary Source Candidates
These videos were discovered outside the official AI Engineer channel and matched by the external-video discovery tool. Treat them as secondary sources only until manually verified against the official event recording.

- [AI Engineer World's Fair: Building Reelful - Agentic Video Editor](https://www.youtube.com/watch?v=AheG9p_JXVw)
  - Uploader: Kate | AI Founder · Building Reelful
  - Confidence label: high
  - Transcript status: captions_imported `raw/sources/external-youtube-transcripts/AheG9p_JXVw.txt`

## Attendance Visibility
No high-confidence attendance icon signal is shown for this talk. The sampled video evidence was either low confidence, source-proxy-only, or did not expose a clear audience view.

## Secondary Interview Context
### Interview Opening (Secondary Source)
Hi, we're building real and we're trying to tackle agenda video editing problem from video editing problem from the agent standpoint. I'm K. I'm founder and co. But let's first talk about what's agentic video editing is.

### Official Schedule Identity
- [[ekaterina-deyneka|Ekaterina Deyneka]] — Founder & CEO at [[reelful|Reelful]].

### Topics Mentioned In The Interview
- [[agentic-search]]
- [[ai-sandboxes]]
- [[coding-agents]]

### Secondary Source Material
- [[youtube-AheG9p_JXVw-transcript]] — transcript markdown; source cache `raw/sources/external-youtube-transcripts/AheG9p_JXVw.txt` (1,340 words).
- [[youtube-AheG9p_JXVw]] — related YouTube source page.
- [[youtube-AheG9p_JXVw-slides]] — slide evidence.

### Interview Concepts / Methods
- No highlighted novel concept has been detected yet.

### Evidence Boundary
This section presents a third-party interview, not an official recording or transcript of this scheduled session. It is retained as secondary context only; the official schedule remains the canonical source for the session title, time, room, status, and speaker attribution.