AutoQC

There's a specific kind of dread every editor knows.

You've just finished a 10-minute brand film for a client like Google. Deadline in 2 hours. And now someone has to sit down and scrub through every single frame — checking for black frames, audio clipping, loudness violations, typos. Frame. By. Frame.

That someone was the QC specialist next door. Or the producer down the hall. Not me — I was the design & editor. But I watched it happen on every single project.

In Germany, there's a word for people who are unnervingly good at catching errors: Adleraugen. Eagle eyes. I never had them. But I watched colleagues burn through hours developing them — and heard exactly what they thought about it during coffee breaks.

After 8 years building commercial pipelines for Google, Apple, Nike, Adidas, I knew this problem wasn't going away. It was just being absorbed as "the cost of doing post-production."

So I built the tool I always wished existed — for them, and for everyone who's ever submitted a video and held their breath.

What I got wrong first

My first instinct was to go all-in on AI. Computer vision for everything. Three months in, I scrapped it. AI was too slow, too expensive, and introduced more false positives than a junior editor having a bad day.

The pivot: Basic video signal processing technique as the foundation, AI only where it genuinely adds value (For example, OCR for subtitle extraction). Boring infrastructure, precise results.

That decision brought Bad Edit detection to 100% precision and Subtitle typo detection to 1% false positives on my test set.

Where I am now

• Launched March, 2026 at autoqc.app

• 14 detection checks: Bad edits, typo detection, speech clarity analysis, platform compliance for YouTube / TikTok / Instagram / Broadcast, Aspect Ratio Detection, Border Detection, Brightness Detection, Blur Detection, Background Noise Analysis, Silence Detection, Safe Zone Check, Subtitle Refinement.

• The workflow is simple: upload your video, go grab a coffee, catch up with a colleague. Come back to a QC report that's already caught ~90% of the issues a manual reviewer would find — timestamped, explained, ready to act on

• Currently in early access.

The thing I'm still figuring out

My background is post-production, not SaaS distribution. I know exactly why a video fails quality control. I have no idea yet whether content creators or post-production houses will convert better as paying users.

If you've shipped a tool for video professionals or creative workflows — how did you find your first real customers? Not signups. Paying customers.

I'm looking for beta testers. Specifically: editors, motion designers, or anyone who ships video regularly and has felt this pain. No pitch — I just want to watch real people use it and find out where it breaks. If that's you, drop a comment or DM me. Free access, honest feedback welcome.