How shivajaysaxena/infosysspringboard works
Project type and purpose
• This is a full-stack, single-page Streamlit application called “Real-Time AI Sales Call Assistant.”
• Purpose: capture live audio from a user’s microphone, run real-time speech-to-text, sentiment analysis, contextual product/spare-parts recommendations, analytics dashboard, and auto-generated reports.
File structure
Top-level files:
• app.py → Streamlit entrypoint / UI layer
• assistant.py → Core AI business logic (calls to Google Gemini, FAISS search, Pandas)
• requirements.txt / packages.txt → Dependencies
• phone_comparison.csv, spare_parts.csv → Data stores for recommendations
• uploads/ → Temporary storage of uploaded or streamed audio files
README for insights
• Tech stack: Streamlit (UI), SpeechRecognition (voice capture), Google Gemini AI (NLP), Pandas + FAISS (data & vector search), Plotly (charts), FPDF (PDF reports).