Real-time, multimodal deepfake detection and watermarking for enterprise
The Hook
“Introducing Agent Detection from Resemble. The first System One model built to detect agents.”
Problem Agitation
“Your website has new visitors. Claude Code, Codex, Pi, and thousands more.”
Problem Agitation
“They were built for APIs, not your pages. So they squint at screenshots, scroll too far, and click the wrong thing.”
Problem Agitation
“And your analytics counts every one of them as a person.”
Product Reveal
“Add one line of script, and Agent Detection starts watching the door.”
Feature Teaser
“One read, and your site knows: person or agent, Claude Code or Codex, the moment it arrives.”
Feature Teaser
“So send it where it works best. Your MCP server, your LLMS.txt, your agent card. The booking that took 40 seconds now takes one.”
Feature Teaser
“And the ones you don't want? Scrapers and credential stuffers are stopped at the door.”
Feature Teaser
“Every visit, person or agent, lands on your Resemble dashboard.”
Call to Action
“Resemble Agent Detection. Know who's at your door. Starting at just one dollar per thousand visitors.”
Synthesizable in Remotion · key components of the agent-detection-launch template.
Multiple distinct cursor pointers move autonomously across website elements displaying custom labels and statuses.
Use useCurrentFrame and interpolate to animate multiple SVG cursor nodes along separate bezier paths with label pills.
Integer metrics rapidly count upward with tabular numeric spacing.
Interpolate frame values to Math.floor integers formatted with Intl.NumberFormat.
A vertical glowing red line acts as a collision boundary reflecting unwanted agent cursors.
Render an animated absolute div with red drop-shadow and trigger bounce transforms on colliding cursors.
Purpose. Establish that automated agents represent a huge unnoticed share of website traffic.
Execution. Surging counter overlaying a live travel website mockup filled with agent cursors.
Purpose. Show the inefficiencies and analytics pollution caused by agents navigating human web UIs.
Execution. Visualizing misclicks, overshoot scrolling, and corrupted analytics graphs.
Purpose. Introduce simple installation, automated routing, and defensive filtering.
Execution. Single-line script integration leading to split routing paths and malicious bot barriers.
Purpose. Summarize value proposition with clear pricing and brand call to action.
Execution. Clean branded end screen highlighting $1/1k visitors pricing and logo mascot.