I’ve gotten to the point where I can supply an XML feed of dynamic AIML.  But, what to do with it?

Is there any cloud based platform with an API able to receive and store such dynamic AIML, and with an interpreter API that could access it in near realtime??

If nothing like this exists as a service, what are the alternatives?  What would a theoretical configuration look like?

It probably depends on your definition of ‘dynamic AIML’.
I suppose you might consider Pandorabots a cloud based server for dynamic AIML.

Basically, I’m now able to convert web feeds into AIML with semantic APIs.  My questions are now, how to “store” this dynamic AIML in such a way that it’s available to an interpreter API with the least amount of latency, as well as how to accomplish this with the least hardware and programming overhead??  The idea is to create “ephemeral” AIML for realtime programmatic interpretation.

I don’t know that you will be able to find any of the AIML servers that would currently do what you want as is. For most, you could upload an AIML file on a regularly timed basis. That might meet your needs. If you want something more custom, you might try adapting RiverScript Python. If I remember right, it can import AIML files.

You might be better off though going straight from RSS to the internal interpreter language.

Bots like Skynet-AI can open an RSS feed directly:
open http://www.npr.org/rss/rss.php?id=1001

or ask:
What’s new?

But, I assume you would like the bot to read/internalize the resulting feed.

Marcus,

The beauty of dynamic realtime may be in not having to store it statically.
Generate it, use it, dispose of it… And when you need it again repeat the
process… Generate it, use it, dispose of it…

Congratulations!

“On the fly” that’s the term I was looking for.
Find an application that may benefit from that
and you may find an answer for what to do
with it.