Document ingestion support for PDF and docx and WAV chunk decoding

[suryam789]

No description provided.

[Copilot]

Pull request overview

Adds richer ingestion and observability features to the smart kiosk stack by expanding RAG context upload to support document formats (PDF/DOCX), enhancing query-time metrics streaming, and improving browser audio streaming compatibility with WAV-container chunks. Also introduces endpoints to serve synthesized response audio segments and to proxy platform hardware info.

Changes:

• Add PDF/DOCX parsing for /api/v1/context/file uploads and batch ingestion response types.
• Extend RAG query request flags and SSE output to optionally include performance/LLM metrics.
• Decode browser-uploaded WAV chunks into PCM frames and add endpoints for response-audio playback and platform-info proxying.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
smart-kiosk-assistant/rag-service/requirements.txt	Adds dependencies for DOCX/PDF ingestion (python-docx, pypdf).
smart-kiosk-assistant/rag-service/dto/query_dto.py	Adds request flags for metrics and new DTOs for per-file + batch ingest responses.
smart-kiosk-assistant/rag-service/api/custom_endpoints.py	Implements DOCX/PDF text extraction, batch file ingestion, and optional SSE metrics emission.
smart-kiosk-assistant/main.py	Adds endpoint to serve synthesized response-audio WAV segments.
smart-kiosk-assistant/kiosk_core/service.py	Adds lookup helper for synthesized response-audio segment file paths.
smart-kiosk-assistant/kiosk_core/audio_session.py	Decodes WAV-container chunks to PCM to stabilize downstream RMS/VAD/ASR processing.
smart-kiosk-assistant/kiosk_core/api/endpoints.py	Adds /api/v1/platform-info proxy endpoint to metrics-collector.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[suryam789]

I have included the changes suggested by copilot.