We are looking for a skilled AI Automation Engineer to build an intelligent system that analyzes video content and performance metrics to provide actionable insights for creators.
The goal is to automate the feedback loop: our system should "watch" a video, "read" its analytics data and comments, compare this with our internal "Winning Video Knowledge Base", and generate precise recommendations on how to improve the video.
Key Responsibilities:
Data Integration: Build a pipeline to ingest video files and performance data (views, CTR, retention curves, drop-off timestamps, comments etc.) from platforms.
Multimodal Analysis: Implement AI models to analyze video frames, audio (transcripts), and pacing to understand what happens at specific timestamps.
RAG Implementation: Develop a Retrieval-Augmented Generation (RAG) system based on our proprietary documents (PDFs, Notion pages, etc.) about viral video structures and storytelling.
Insight Generation: Create a logic where the AI identifies why viewers dropped off at "Second 45" by correlating the retention dip with the visual/audio content and comparing it against best practices.
Automation Workflow: Design a seamless workflow (n8n, Make) that outputs a structured "Improvement Report."
Technical Requirements:
LLMs: Deep experience with OpenAI (GPT-4o), Claude 3.5, or Gemini 1.5 Pro.
Computer Vision: Experience with video processing or using multimodal models for video understanding.
Database: Experience with Vector Databases (Pinecone, Weaviate, or ChromaDB) for the knowledge base.
Data Analysis: Ability to work with APIs and process JSON/CSV data from analytics dashboards.
To be considered, you must provide:
A similar AI automation project you’ve built.
Which tools/models you would use for this project.
Apply Now
Apply Now