Inspiration
Creation of fully automated system that detects viral trends early and turns them into monetized short videos faster than anyone else.
What it does
AutoViral finds new social media trends, generates videos with subtitles and audio, posts them automatically, tracks performance, and boosts successful ones — all under human control via API.
How we built it
Using Node.js (Express API), BrowserUse automation, ffmpeg, Daytona sandboxes, multiple LLMs (OpenAI, Gemini, Claude), and royalty-free media from Pexels. Typescript for front-end and AI studio. (cmon bro we had only ~5 hours)
Challenges we ran into
Balancing speed and safety, managing ephemeral sandboxes, API rate limits, and coordinating multiple LLM and browser automation workflows.
Accomplishments that we're proud of
A working end-to-end pipeline: from trend discovery to posting and metrics tracking — all automated and deployable with one script.
What we learned
A lot! (yes)
What's next for AutoViral
Implement multi-armed bandit optimization, improve style per platform, add real-time dashboards, and introduce human approval modes for brands.
Built With
- browseruse
- daytona
- ts
Log in or sign up for Devpost to join the conversation.