Vidzly
Team consisting of Viet‑Tien Ngoc (CTO, Sky Solution; Go/Python/TypeScript; Vietnamese STT models), Hanh Thi‑Hong Tran (NLP Engineer, Arkhn; Python/LLMs), and Linh (PhD, UVSQ; 5G/security, ROS).
YouTube Video
Project Description
Vidzly - Your brain designs it. Our AI edits it. You hit publish.
This project, Vidzly, is an Autonomous Influencer Pipeline built during the hackathon to solve the critical friction point for aspiring micro-influencers: the need for advanced editing skills and equipment to produce viral short-form content.
- Core Problem & Target Customer
The core problem is the high barrier to entry in short-form video (TikTok, Reels, Shorts) caused by complex, time-consuming post-production. Our target customer is the beginner micro-influencer (the “zero-skill, zero-equipment” creator) who has great content ideas but lacks the ability to edit effectively.
- Main Features & Demo Flow
Vidzly automates the entire post-production process:
- Input: The user provides a Creative Brief (Vibe) and a Raw Video Clip.
- Flow
- AI Analysis (Gemini + ElevenLabs): Gemini analyzes the brief and ElevenLabs transcribes the raw video, identifying the best up to 30-second segment, key emotional moments, and metadata requirements.
- Autonomous Workflow (n8n): The n8n orchestrator generates a comprehensive editing blueprint.
- Lovable to make our front-end come true with strong vibe coding.
- Output: The final, polished up to 30-second video is rendered with thumbnail is prepared for multi-platform.