Looking for an AI Video + FFmpeg Expert to Build Conversational Video Editing Solution We are building an AI-powered video editing system where users can edit videos simply by talking to the AI — no manual timeline editing. Your role is to architect and develop this system using LLMs + automated FFmpeg pipelines. Core Requirements: Build a system that understands natural language commands such as: “Add this image as an overlay between 00:03–00:06” “Apply a slide transition between scene 1 and scene 2” “Replace background music from 10 seconds onwards” “Cut this part and zoom in on the product” Convert user instructions into structured edit actions. Programmatically generate and execute FFmpeg filter graphs to apply overlays, transitions, effects, cuts, audio changes, etc. Support multi-step conversational editing (user can refine: “Make the overlay smaller”, “Slow the transition”, etc.). Build preview and final-render pipelines (low-res preview + full-quality export). Integrate scene detection, overlays, and other visual cues as needed. Ideal Candidate: Strong experience with FFmpeg (complex filter graphs, transitions, overlays, timeline-based edits). Solid understanding of AI/LLM workflows — instruction parsing, prompt engineering, structured output, etc. Experience building automated video pipelines (Node.js/Python). Ability to architect end-to-end systems (analysis → edit plan → render). Bonus: experience with scene detection, computer vision, or ASR. We are building a next-gen “Talk-to-Edit” AI video editor. If you can translate human instructions into automated FFmpeg operations, we want to work with you.