Smart Customer Support Chatbot Development
Eira Gozum
Intent recognition: orders, refunds, FAQs, escalation.
Human fallback, logging, analytics.
Modular code, Docker deployment, and best practice structure.
Get status by order ID.
Automated refund/return policy info.
Query a knowledge base using csv data (LangChain supported).
Human agent intent/escalation.
Smart fallback (OpenAI/Gemini/HF) for unknown/complex queries.
Track usage, trends, session logs.
Easy Docker Compose or local setup.
Rasa 3.x (core NLP, dialogue)
Python 3.9+ (all custom logic)
FastAPI (integration/backend)
Streamlit (frontend, dashboard)
MongoDB/Redis (DB, cache)
Hugging Face/OpenAI API (fallback/LLM)
Docker, Docker Compose (infra)
Then (in separate terminals):
Hello! How can I help you today?
: “hi”, “hello”, “hey”
: “track my order”, “where is order #12345”
: “what is your refund/return policy?”
: “how to reset password?” etc. (FAQs via CSV)
: “I want to talk to a human”
: “bye”, “good night”
: unknown, gibberish, complex
: “are you a bot?”
/tests/test_stories.yml
Custom test convos: update data/sample_conversations.csv as you want.
Easiest for production/local demo.
Cloud (Render, etc):
Default ports - Rasa (5005), Backend API (8000), Frontend (8501), Analytics (8502).
Like this project
Posted Aug 17, 2026
Developed a smart customer support chatbot with intent recognition and analytics.
Likes
0
Views
2
