How kawtherah/project2-reception-robot works
Project type and purpose
• This is a hybrid chatbot project (“Reception Robot”) combining a static web front-end (Panel.html + ss.css) and a Python-based voice interface module (watson-streaming-stt) that together interact with IBM Watson cloud services (Assistant, Speech-to-Text, Text-to-Speech).
• The overall goal is to receive customer queries (via text or spoken language), interpret intent/entities via Watson Assistant, respond via text on the web panel or synthesized speech, and log customer satisfaction.
File structure
Top-level:
• Panel.html / ss.css – the static web UI for the chatbot
• skill-Recommendation-Chatbot-skill.json – Assistant skill configuration (intents/entities/dialog)
• README.md – high-level project description
watson-streaming-stt/ (voice module):
• transcribe.py – Python script orchestrating STT → Assistant → TTS
• speech.cfg – credentials & endpoint config for Watson STT/TTS
• chatpotResponse.txt, UserInput.txt, OutputAsSpeech.mp3 – I/O artifacts for prototyping
• requirements.txt, setup.py – Python package & dependency config
README for additional insights
• The README highlights three Watson Assistant components (Intent, Entities, Dialog) and a separate “Voice Chatbot” task using IBM Watson services with Python.
• There is a deployed static web version (GitHub Pages) and a downloadable skill JSON.
Main components & interactions
a. Components
1) Web Front-End: Panel.html + ss.css (+ JavaScript logic to call APIs)
2) Voice Backend Service: Python module (transcribe.py)
3) IBM Watson Assistant service (cloud)
4) IBM Watson Speech-to-Text service (cloud)
5) IBM Watson Text-to-Speech service (cloud)
6) Optional Local File Store (UserInput.txt, chatpotResponse.txt, OutputAsSpeech.mp3) for prototyping/logging
b. Relationships / Data flows
• Text Chat Flow:
User types → Front-End JS → Watson Assistant API → Assistant returns JSON → Front-End renders response
• Voice Chat Flow:
User speaks → Python transcribe.py streams audio → Watson STT → transcripts written to UserInput.txt → Python sends transcript to Watson Assistant → assistant JSON response → Python calls Watson TTS → OutputAsSpeech.mp3 → front-end or local player plays audio
• In both flows, the Assistant skill (skill-Recommendation-Chatbot-skill.json) drives intent/entity extraction and dialog decisions.
c. Architectural patterns / design principles
• Separation of concerns: static UI vs. voice-processing backend
• API-driven microservice interaction: Python service and front-end each call IBM Cloud APIs
• Config-driven connectors: speech.cfg & setup.py manage credentials and dependencies
d. Key technologies
• Front-end: HTML, CSS, JavaScript (fetch or SDK for Assistant)
• Backend: Python 3.x, requests or IBM Watson SDK for Python
• Cloud services: IBM Watson Assistant, Speech-to-Text, Text-to-Speech
Elements
• Front-End (blue), Backend service (green), External cloud services (orange), Data stores/logs (gray)
• Annotations for configs (speech.cfg, skill JSON) attached to respective components