Skip to content

AI Class Project - Building AI agent with effective conversation

Notifications You must be signed in to change notification settings

dunky-star/Intelligent-Agent-Chatbots

Repository files navigation

Intelligent-Agent-Chatbots

[Harnessing the power of LLM - Prompt Engineering]

[ChatBots] why are they the next big thing?
“Conversation as a Platform” - Better customer service and they could even differentiate your business from the crowd. 
Examples: Apple – Siri, Microsoft – Cortana, Google – Google Assistant, Amazon – Alexa, and Meta “Facebook” – ITS ALIVE.io, to name but a few.

A ChatBot simulates human-to-human conversations with automated responses thus replacing the need for an actual human representative,
thus, cutting a lot of costs and creating efficiency. Users in both B2C (Business-to-Consumers) and B2B (Business-to-Business) 
are still using FAQs and raw-based Bots, now, what if we replace these with an intelligent agent utilizing the power of Deductive 
reasoning using both KB (Knowledge-Based) on actual customer conversations in the past and Indcutive learning? 
Harnessing the inductive learning capabilities of NLP through the power of deep learning that allows agents to more or 
less learn and improve based on experience. These give a personalized feel to customers and better accuracy.

The winning team:

1. Geoffrey Duncan Opiyo
2. Phuong Khanh Nguyen
3. Deo Mugabe
4. Duc Phi Ngo

How to run the program (packages to install)

1. python -m venv dunky  -> To create virtual environment
2. cd /dunky -> Change directory to the newly created virtual envornment
3. Scripts/activate -> To activate virtual environment in Windows
4. source/activate -> To activate virtual environment on MacBook
5. pip install "openai<1.0.0" langchain
6. pip install chromadb
7. Create a .env file inside the root directory of the project
7. Get OpenAI api_key, put it inside .env file [file content]: OPENAI_API_KEY= <your key>