As I see this project contain 4 big parts:
1. Realtime processing of network stream
2. Detecting RTP and SIP traffic, catching sessions
3. Saving RTP flow according sessions. Saving SIP according sessions.
4. Converting RTP from G.711, G.729A, G.723.1, G.722, and GSM6.10 to some other format.
As I see hardest part is catching sessions, so we could successfully save flow from different calls.
I previously work with VOIP software for callcenter, developed all routes, make a lot of work with resolving issues so I have enough knowledge to catch traffic and read SIP/RTP protocols.
I suppose this project will take at least one month, but probably will take two.
Also I have big price for my time so you must be ready dramatically increase your budget.
PS: Please specify which one codec did you want to use when saving to wav container.