I am looking for an experienced Python Developer to build a specialized tool for automating video editing workflows. The goal is to create a desktop application (macOS) that segments voiceovers, generates semantic tags for footage, and exports a ready-to-use Adobe Premiere Pro XML project. The Workflow: Audio Alignment: Take a pre-recorded voiceover and a text script, then perform forced alignment (word-level timestamps) using stable-ts or OpenAI Whisper. AI Semantic Tagging: Use Gemini 1.5 Flash API to analyze text segments and generate visual tags based on a "Golden Standard" (a reference database of 50 successful video hooks). Vector Search (The "Spotify" Method): Match the generated tags with a local library of video footage using Vector Embeddings and Cosine Similarity (ChromaDB or similar). Logic & Penalties: Implement a system to avoid footage repetition and ensure high-paced visual variety. Premiere Pro Export: Generate a Final Cut Pro XML (compatible with Premiere) that includes the voiceover and all matched footage cut to the correct timestamps. GUI: A clean, minimalist desktop interface (CustomTkinter or Streamlit). Key Requirements: Expertise in Python (Asyncio is a must for performance). Deep experience with FFmpeg and audio/video processing. Experience with LLMs (Gemini/OpenAI) and Prompt Engineering (RAG / Few-shot prompting). Knowledge of Vector Databases (ChromaDB, Pinecone, or FAISS). Experience generating XML/EDL for video editors. Ability to optimize for speed (Goal: 2 mins processing for a 1.5-min video). Timeline: 2 weeks (Sprint). To apply, please answer: Have you ever worked with Premiere Pro XML or EDL generation? Which tool would you use for forced alignment of audio and text? How would you handle a vector search for 1000+ video files locally?