You can certainly extend Splunk with custom visualizations from whatever librarys you wish to integrate (d3, rickshaw, three.js , google charts etc...), or maybe a commercial library or something you've written yourself.
The approaches I would consider :
1) Create a custom advanced XML module that you could then declare inside of your Splunk UI app form or view XML. Here's a good example of such an approach on SplunkBase.
2) Create a standalone web app that uses the Splunk Javascript SDK from the browser to communicate with SplunkD (via a proxy to get around CORS issues) , and code the custom visualization logic in Javascript in the webpage sent to the browser.
3) Use the any of our language SDKs in a web server to communicate with SplunkD , and code your custom UI javascript in the webpage sent to the browser.Comms between the browser and the web server might use web sockets even.
4) Use the new Application Framework. to achieve options 2 and 3 in a much more rapid and integrated fashion , leveraging your Javascript and Python skills. Here is a tutorial for adding a custom D3 visualization with the Application Framework.
