Twitter API grabbing/caching/parsing multiple streams

I need some thoughts in this :) I think i might be trying to do the impossible.

I’ve got a website with a lot of company profiles (300+).

I want to place a stream on every profile page, and allow the owner of that profile page to enter some keywords in the backend with the keywords they want the stream to display.

1 thing would be to do this with javascript on the fly, on every profile page.

But, since I want to use the combination of every profile stream on the frontpage and parse images in a tweet (yfrog, plixi) i thought it would be best to cache them into a mysql db, and work from there with PHP. (Maybe SEO advantage over Javascript? )

Is this even possible ?

My first bet was to make a REST call to the api with the keywords from all company profiles ,put t开发者_开发百科hem to the DB and store which keyword would be connected to that tweet.

Any thoughts on this ?

You could use the Streaming API with a track filter, that would allow you to capture all statuses matching the keywords your profile owners select. By default, you're limited to 400 keywords, but could ask Twitter to increase you're access level (they list a few levels in the documentation).