Our customers have millions of users who frequent their platforms on a daily basis. There is a fine line between showing the user the most relevant content and boring the user with the same results. How do we show the best results while still keeping the content looking fresh?

Serving one-to-one curated content for each of the millions of users that reach the platform. This is a multifaceted problem. How do you personalize the content, last video seen, last search, most liked videos, and similar videos liked by other users? Even if you solve the “what to show” problem, keeping and accessing millions of different profiles a day without skipping a beat on the video play functionality is no small feat.

We all use Youtube and other streaming video sites. Don’t you want to write the tech that will optimize the image of each of the millions of thumbnails available to view at every moment? Make the thumb too attractive and its clickbait, too boring and no one will click on it. What is the right recipe?

Ranking content is the holy grail of the internet. Google does it all day and night. In our case, we have data that google does not. Make use of that data to rank content based on how it was viewed, past history, query time and more. Think you have the skills to improve our ranking engine, come and give it a try.

Synonyms, misspellings, alternate uses and compound text structures. What did the user mean when they typed something in the query box. Many searches are not obvious, is it a misspelling or the name of a celebrity or slang for a verb. Stitching together an NLP engine that continuously states up to date is a necessity if we want our platform to be understood.

How to use live data to alter future results. What if content is ranked high and the user scrolls past it and clicks on something ranked lower. Should the selected content be boosted? How about the highly ranked content, should it be pushed down?

With the use of image recognition technology its possible to automate the categorization of the video, as well as tagging specific elements in the video. Does it take place on a beach, or in the forest, is there a sunset?

Is there something missing? Are you working on something Next Generation that we are not aware of. We have an eternal hunger for new technologies.