🔥Build an Autonomous AI Video QA / Editor Agent

We are looking for an experienced AI developer to build a standalone Video QA Editor Agent that can plug into our existing software through an API. ✅Workflow Our software sends the agent: 💠AI-generated video 💠Original video prompt 💠Script/dialogue 💠Scene beats and expected actions 💠The agent reviews and analyzes the video against what was supposed to happen. 💠 identifies visual problems, AI anomalies, continuity issues, incorrect actions, framing problems, text errors, etc. ◾Examples: VERY IMPORTANT USES CASES THAT THE AI QA EDITOR WILL CATCH AND UPDATE 🔸Person says “1, 2, 3” but fingers show the wrong count. 🔸Person points at a company logo, but the logo is missing. Talking-head shot accidentally frames below the person's waist. 🔸Person drinks while supposedly speaking. 🔸Misspelled text appears on a sign or product. 🔸Hands, faces, objects, or environments change unnaturally. 🔸The agent outputs: Problems found Time/location of each issue Severity Updated generation prompt designed to correct those problems We then send the revised prompt back through our AI video-generation system and can repeat the QA process. ◾The system should function as an autonomous visual video editor / quality-control agent, not simply a video summarizer. You will also help evaluate available multimodal AI/vision models to find the best balance between accuracy, quality, speed, and API cost. ◾This must be built as a modular API-based component that we can plug into a larger software platform. ✅To Apply Step 1️⃣: Send 1–2 links, demos, or screenshots of impressive software you have personally built using AI as an agent or autonomous system. Briefly explain what each system does. Step 2️⃣: Explain your experience building autonomous AI vision/multimodal agents, especially systems that analyze images or video and make decisions based on what they see. Step 3️⃣: Confirm that you are available to begin immediately and give us your approximate estimate for how long you believe a working first version of this module would take to build. We look forward to working with you!