ScreenCI
ScreenCI generates narrated tutorial videos of a real application from a single text prompt, which users paste into any coding agent. The agent records a walkthrough with auto-zooming camera tracking, narration, subtitles, and multi-language support, and videos can be re-recorded automatically on every code push via CI integration. The browser-based editor lets non-engineers update narration, voices, and overlays without touching code.
- Developer Tools
- Artificial Intelligence
- Software Only
Funding
Founders
Founder details are not available yet.
Product
Problem
Creating product tutorial videos is a slow, manual process that typically requires video editing tools, screen recording software, and technical knowledge. Teams often resort to outdated mockups or static screenshots because recording real application flows takes significant engineering time, and keeping videos current as the product evolves becomes an ongoing maintenance burden.
Solution
ScreenCI generates narrated tutorial videos of an application from a single text prompt. Users describe a feature or desired change in plain language, paste a generated prompt into any coding agent, and the agent records a walkthrough of the real running application—not a mockup—with professional production quality. The output includes auto-zooming camera tracking that follows every click, field entry, and drag, plus narration, subtitles, branding, and multi-language support. Videos are rendered for help centers, documentation, social media, or in-product placement, and scripts can be edited later in a browser-based editor without prompting an agent again. The recording runs locally on the user's machine or CI, so source code and internal environments never reach ScreenCI's servers.
Target Audience
Product teams—including engineers, product marketing, support, and documentation contributors—who need to produce and maintain up-to-date tutorial videos for their software applications.
Features
- Auto-zoom camera system that dynamically pans and zooms to follow user interactions, including typing, segmented controls, slider drags, and dropdown selections
- Works with any coding agent—Claude Code, Cursor, Codex, desktop agents, or custom harnesses—that can run a shell command and access Node.js
- Adds one or multiple narration languages to any video, automatically translating and recording the script (up to 79 languages)
- Media overlay support for embedding PNG/SVG images and MP4 clips as timed assets on the video timeline
- CI integration (Add to CI) that stores a credential in the user's CI provider and records videos automatically on every push, preventing script drift
- Browser-based editor enables non-engineers in product marketing, support, docs, and engineering to edit narration, voices, overlays, and languages without opening the video's code
- Open-source CLI and recorder for auditability of what runs against the user's application