S2S: Speech-to-Speech with File Search in a single API Call

[Prajna1999]

Is your feature request related to a problem? Please describe.
Users have a need to take in voice notes (in indic languages or english) as questions, and generate answers based on their knowledge base + prompt and provide them back as voice notes (in same language as the voice notes). Currently the user need to take care of three sandwich API calls (stt -->rag -->tts) for each s2s call.

Describe the solution you'd like
One single API endpoint with minimal config to do s2s API call.

Additional context
Glific S2S requirements
S2S PRD