perf: implement benchmark-driven optimizations

fix(rag): remove english translation step & increase top-k chunks

feat(rag): switch to multilingual embedding model

feat: auto-load PDFs, CSVs, and TXT files from data/ directory

fix: robust translation (3-tier fallback), dual-query RAG, pipeline safeguards

fix: use gemini-2.0-flash, robust Hinglish translation, RAG k=5

Switch to Mistral-7B API: clean prompts, fix UI labels, restore HF_TOKEN warning

Clean rewrite: whisper-small local, clean RAG+LLM with Hindi prompts

Migrate RAG from LangChain to LlamaIndex

Add BankFAQs dataset and fix langchain-community dependency

Clean commit without venv