dating site powered by ai tagging, made by ai
Budget: $250 – $750 USD
I’m kicking off a new video-centric platform and the very first milestone is a rock-solid photo & video publishing module that can tag every upload automatically. Instead of static, user-entered keywords, I want AI-generated tags produced through both image recognition and full video content analysis so that clips and photos instantly become searchable the moment they finish uploading.
Here’s what I need built right now:
• An upload pipeline that accepts large videos as well as photos, performs any basic transcoding/compression required, and hands the media to the tagging engine.
• An AI tagging service that runs frame-level image recognition and video content analysis, returns a ranked list of tags with confidence scores, and stores them in a database I can query later.
• A lightweight interface (web or API-only is fine for now) where I can review, approve, or override the suggested tags before the asset goes live.
• Clear documentation and setup scripts so I can reproduce the entire stack on my own server or a cloud instance.
Acceptance criteria: when I upload a test video or image, the system should deliver at least five accurate, context-appropriate tags with ≥90 % confidence, all within a reasonable processing time.
Preferred tech is flexible—Python, Node.js, TensorFlow, OpenCV, AWS Rekognition, Google Vision, or similar are all on the table—as long as the solution is modular enough to plug into future features such as random 1-to-1 live cams and casino-game integrations.
If you’ve built anything similar or have relevant demo links, I’m eager to see them. Let’s get this tagging engine up and running!
Here’s what I need built right now:
• An upload pipeline that accepts large videos as well as photos, performs any basic transcoding/compression required, and hands the media to the tagging engine.
• An AI tagging service that runs frame-level image recognition and video content analysis, returns a ranked list of tags with confidence scores, and stores them in a database I can query later.
• A lightweight interface (web or API-only is fine for now) where I can review, approve, or override the suggested tags before the asset goes live.
• Clear documentation and setup scripts so I can reproduce the entire stack on my own server or a cloud instance.
Acceptance criteria: when I upload a test video or image, the system should deliver at least five accurate, context-appropriate tags with ≥90 % confidence, all within a reasonable processing time.
Preferred tech is flexible—Python, Node.js, TensorFlow, OpenCV, AWS Rekognition, Google Vision, or similar are all on the table—as long as the solution is modular enough to plug into future features such as random 1-to-1 live cams and casino-game integrations.
If you’ve built anything similar or have relevant demo links, I’m eager to see them. Let’s get this tagging engine up and running!