AI Developer Needed to Build Modular YouTube Video Monetization Scanner - Contract to Hire

I'm building a tool that helps YouTubers scan their videos before uploading to determine if their content may violate YouTube’s monetization policies. The system should analyze visual, audio, and textual content to flag potential issues such as profanity, nudity, violence, copyright risks, etc. Important: The system should be modular so it can start using third-party APIs (e.g. Google, OpenAI) but later swap to open-source models (e.g. Whisper, YOLOv8, CLIP, NudeNet) with minimal rework. Responsibilities: Build a backend system that accepts video input (MP4, etc.) Use AI tools to: Transcribe audio and detect profanity/hate speech Analyze video frames for nudity, violence, blood, drugs, weapons Scan text overlays or subtitles for flagged words Return a report that shows where issues were found and why (timestamped) Build it so I can plug it into my Wix website (or deploy via API/backend) Build it modularly so each analysis area (audio, vision, etc.) can be easily switched between API and open-source models later Provide brief documentation for future devs Tech Stack Preference (flexible): Python for backend Flask or FastAPI (if web-based) Google Video Intelligence API, Whisper API, OpenAI’s Moderation API (initially) Later use open-source (Whisper (local), YOLOv8, NudeNet, Detoxify, CLIP) OpenCV / ffmpeg for frame extraction if needed Ideal Candidate - Experience with video/audio AI analysis Comfortable using third-party APIs AND open-source models Knows how to build modular codebases Has worked on AI moderation tools, media analysis, or YouTube-related projects Milestones & Payment: This is a milestone-based job. Milestone 1: Planning & Architecture Design — $100 Define exact tools for each module (APIs now, open-source later) Show flowchart of modular system architecture Milestone 2: MVP (API-based version) — $300 Basic working scanner using third-party APIs Upload video → get report with issues flagged Transcription + vision + basic profanity flagging Milestone 3: Wix/API Integration Ready Version — $150 Scanner deployable to work from Wix or via simple backend endpoint Polish and test (Optional Later): Open-source Module Integration — To be discussed Replace API modules with open-source alternatives To Apply: Please reply with - Examples of AI tools you’ve built (especially video/audio/moderation tools) What stack you’d recommend for this Whether you’re comfortable with both third-party and open-source integrations No strict deadlines — we prioritize quality over speed. Flexible with timezones. Bonus available for high-quality Apply Job!