Why I Built This
We search. We go back. We verify.
The UI for this just didn't exist.
The Frustration
I hated the 1D seek bar. Scrubbing back and forth, guessing where that scene was, wasting time on trial and error. Videos aren't linear experiences anymore. Why should navigation be?
The First Attempt
At first, I just spread the video into a grid of thumbnails. But that wasn't enough. I couldn't tell where I was. The grid was a static map, but video is a temporal experience. They were disconnected.
The Marker
I needed a marker that moves smoothly across the grid, synchronized with playback. The moment I made it touchable, everything changed. The grid became a space you could explore, not just view.
The Final Insight
A single grid resolution misses details. I needed dynamic time granularity - the ability to zoom in and out of time itself. With smart caching, the grid regenerates instantly at any resolution.
The Journey
VAM started as a personal Python project. Over 3 months, it grew into 100,000+ lines of code - a desktop application pushing the limits of what one developer could build. But I wanted to share the core innovation with the web.
The Essence
VAM Seek is the essence of that journey, distilled into ~36KB of JavaScript (with Multi-Video Cache). No server. No dependencies. Just the pure idea: see your video in 2D, navigate instantly.
Grid Seek Marker is not a UI for watching videos.
It's a tool for exploring time.
This is not a new feature.
It's a missing standard.
Once you use it, going back feels incomplete.