"I Like Editing. It's This Part I Can't Stand." — A Real Demo with a Working Editor

Three and a half hours of Bruce Fromong telling stories about his 30-year friendship with OJ Simpson. Sitting on a hard drive. No investors yet. No budget to hire someone else to cut it down.

That's what Mark Leffingwell walked into our demo call with. He and his wife Vicki run a production company, and they need a 3-to-5 minute sizzle reel pulled from that mountain of footage so they can pitch it to investors. But they're also running other client projects. They don't have days to sit and re-watch everything.

"I would rather go get my teeth pulled out of my head with no anesthesia than sit there and listen to three and a half hours of Bruce telling stories again," Mark told me. "I heard the stories. I don't want to go through it again."

And that's the thing nobody talks about in video editing. The creative work is great. Color grading, transitions, building a story. That's the part editors signed up for. The part that burns people out is everything that comes before it: watching hours of interviews you've already sat through once, hunting for that one soundbite you vaguely remember, manually cutting and collecting clips into bins.

Mark put it perfectly: "I don't think the burnout was because of the editing. I like editing. It's this part of it."

What "This Part" Actually Looks Like

Here's the workflow most editors know too well. You shoot a bunch of interviews. You import everything into Resolve or Premiere. Then you sit down and listen to every single interview again, taking notes, marking timecodes, trying to pull out the best soundbites and figure out a structure.

Mark's team had been experimenting with a manual version of AI assistance. They'd transcribe interviews, export the transcripts, paste them into ChatGPT, ask it to help organize the story, then take that output and manually bring it back into DaVinci Resolve. It worked, sort of. But ChatGPT kept rewriting things, mashing sentences together, and the whole process still involved a lot of bouncing between apps.

What Happened When We Ran It Through Vindexr

During the demo, I showed Mark a similar situation I'd been working with. A folder full of interviews about lean manufacturing principles. Hours of material that needed to become a three-minute piece.

Instead of listening through everything, I opened the AI tab, scoped it to the interview folder, and typed a simple prompt: "Create a collection of all the most interesting soundbites arranged in good story structure."

Vindexr's auto-transcription had already processed all the audio during import. So when Claude analyzed the content, it had full transcripts to work with. It came back with 27 selections, each with a summary explaining why it was relevant. The first clip was tagged as a good opening because it defined the core topic. Each subsequent selection had context for where it fit in the narrative.

No re-watching. No manual timecodes. No copy-pasting transcripts between apps.

And the part that really clicked for Mark: I hit "Create Collection," and all 27 clips were arranged in the sidebar. Then I sent the whole thing to Resolve with one click. A new bin appeared with all the referenced media, and a timeline with every cut already placed.

"Okay, that's super handy," Mark said. Then the wheels started turning.

Where It Got Real

Mark's wife Vicki is the story brain on their team. She's the one who goes through transcripts and builds the narrative structure. Mark handles the technical side, the color grading, the finishing work.

He immediately saw how this would fit their actual workflow. Vicki could sit in Vindexr, run different queries against the Bruce Fromong footage, read through what the AI surfaced, rearrange clips, build out collections. Then she could send everything to Resolve for Mark to do the edit.

"She could literally sit through and read what it came up with, make changes, and then export to DaVinci for me to do the other parts."

No need for Vicki to learn Resolve. No need for Mark to sit through three and a half hours of footage before the creative work even starts.

Mark also connected it to their B-roll workflow. Their process is interviews first, then B-roll planned around what comes out of the interviews. With visual search, he could just query things like "motorcycle grip shot" or "barn" or "sunset" and have Vindexr find and collect those clips instantly.

"Instead of having to search all the things, I can just go: where's that motorcycle grip shot? Where's that flower? Where's that sunset? And it just goes, here it is. And then organizes it for me."

The Burnout Problem Nobody's Solving

Something Mark said stuck with me. He's been doing this a while, and he was hitting a wall.

"I was getting burned out, man."

Not from the work he loves. From the work around the work. The hours of scrubbing through footage. The tedious process of collecting and organizing before you can even start making creative decisions.

This is the part of editing that AI can actually help with without stepping on anyone's creative territory. Nobody's worried about AI finding the right timecode for a soundbite. Nobody's worried about AI organizing clips into a bin. That's not creative work. That's logistics.

The creative work, the choices about what story to tell, what moments hit hardest, what order builds the best narrative, that stays with the editor. Vindexr just gets you to that part faster.

What Mark's Testing Next

Mark left the demo planning to import all the Bruce Fromong footage over the weekend and start experimenting with different prompts. His idea was to have Vicki focus on the story queries, things like "find all the emotional, impactful moments" or "what are the highlights from the first 30 minutes," while he handles the technical setup.

The goal: take three and a half hours of interview footage and get to a sizzle reel without either of them having to re-watch everything from scratch. No $5,000 budget needed to hire it out. Just their existing footage, their story instincts, and a tool that handles the tedious middle.

If that Bruce Fromong sizzle reel lands them investors, the tool paid for itself before they even finished the first project.