ruNNtime — ruNNtime — neural nets in the browser on TypeGPU (Product Launch Video Storyboard & Audio Breakdown)

ruNNtime is a library for running neural networks in browsers, built on TypeGPU. The demo runs speech-to-text plus a privacy filter together on-device.

4-Beat Launch Video Storyboard Timeline (8s Total Runtime)

  • Beat 01 · Pattern Interrupt Hook (0%–15%): Terminal → Live UI Canvas Cut — Immediate visual focal lock before logo bumper.
  • Beat 02 · Friction & Problem Contrast (15%–40%): Contrasts legacy workflow bottlenecks against ruNNtime's streamlined interface.
  • Beat 03 · Core Product Proof (40%–80%): High-clarity UI macro capture demonstrating ruNNtime's primary capability in real time.
  • Beat 04 · Payoff & Call to Action (80%–100%): Brand lockup, launch availability, and URL callout.

Audio & SFX Engineering Profile (EBU R128 Telemetry)

Paced at 122 BPM with an integrated loudness of -12.8 LUFS. Sound design layers mechanical keycap clicks, soft UI whooshes, and sub-bass transition drops synchronized to visual scene cuts.

60fps Remotion React Code Specification

import { AbsoluteFill, Sequence, spring, useCurrentFrame, useVideoConfig } from "remotion";

// ruNNtime — ruNNtime — neural nets in the browser on TypeGPU (8s @ 60fps | 122 BPM | Hook: Terminal → Live UI Canvas Cut)
export const LaunchComposition = () => {
  const frame = useCurrentFrame();
  const { fps } = useVideoConfig();
  const cameraZoom = spring({ frame, fps, config: { damping: 16, stiffness: 140, mass: 0.8 } });

  return (
    <AbsoluteFill style={{ backgroundColor: "#141413", transform: `scale(${0.96 + cameraZoom * 0.04})` }}>
      <Sequence from={0} durationInFrames={72} name="Beat01_PatternInterrupt_Terminal → Live UI Canvas Cut">
        {/* 0-15%: Terminal → Live UI Canvas Cut cold open */}
      </Sequence>
      <Sequence from={72} durationInFrames={120} name="Beat02_WorkflowContrast">
        {/* 15-40%: Problem friction vs ruNNtime solution */}
      </Sequence>
      <Sequence from={192} durationInFrames={192} name="Beat03_ProductMacroProof">
        {/* 40-80%: Real-time UI macro capture (runntime) */}
      </Sequence>
      <Sequence from={384} durationInFrames={96} name="Beat04_BrandPayoffCTA">
        {/* 80-100%: Brand lockup + launch CTA */}
      </Sequence>
    </AbsoluteFill>
  );
};

Frequently Asked Questions

How is the ruNNtime product launch video structured?

The ruNNtime launch film runs 8 seconds and follows a 4-beat conversion arc anchored by a Terminal → Live UI Canvas Cut opening.

What tools and motion stack were used?

Detected motion characteristics include 60fps spring-physics camera easing, crisp UI callout zooms, and EBU R128 broadcast audio mastering (-12.8 LUFS).