Hi there!
We’re looking for a programmer and data engineering team to build a real time or near real time (Uber Hudi perhaps) KNN ML model that finds the nearest neighbors of three data sets:
1. audio of streaming media or social media
2. visual scenes / frames / images from streaming media or social media
3. available inventory from eCommerce providers, specifically Google Shopping or Amazon (using their API’s)

The end goal is: A user is watching Netflix. User sees an item they like (book, shirt, pants, shoes, etc.). User pauses netflix. User takes a screenshot of the frame with the image of the desired item (book, shirt, pants, shoes, etc.). User uploads screenshot. App that is being built via this job post provides the desired item – which was matched on the back end via KNN ML models – from the eCommerce metadata platform (AliBaba, Amazon, Google Shopping)

This KNN ML model would also match these data sets for users browsing their social media feed.

This helps provide more background.  Netflix Tech Blog explains how they used KNN and contrastive learning to correctly categorize specific scenes based on audio, visual, and text (labeling data): https://medium.com/netflix-techblog/rebuilding-netflix-video-processing-pipeline-with-microservices-4e5e6310e359

Uber Hudi: https://www.uber.com/blog/hoodie/

Thank you!

Posted On: January 20, 2024 22:56 UTC
Category: Mobile App Development
Skills:Mobile App Development, iOS Development, Android App Development, Android, iOS, Microsoft Windows, Desktop Application, Ecommerce Platform, Machine Learning Model, Instagram

Country: United States

click to apply