Max-Stream
Streamlit Application
Project maintained by Kanka-max Hosted on GitHub Pages — Theme by mattgraham
Jacques Patricks Porfolio
Data scientist & Conversational AI Designer. Connect on LinkedIn or Twitter
Retail Analysis Data App
This demo application provides business analytics and intelligence for a sample retail store. The data stories have been visualized using Plotly charts. This data application is deployed on Streamlit Cloud.
App Functions
- Business Selfie
- Customer Segmentation (Using Clustering Model)
- Prediction on Future Sales (*Using Long-Short Term Memory Neural Network)
Hotel Digital Assistant.
This demo of a hotel digital assisant is a conversational AI use case. The application is built on a Rasa AI Open Source Framework. It is also deployed via telegram external channel. The deployment files were hosted on an Azure Virtual Machine. Continuous integration and continuous deployment is facilitated by github and Rasa X.
App Functions
- The app uses DIET language model and TED Transformers to predict user intents and bot responses.
- Enables the user to make orders for burger and chicken.
- Captures customers details (name and address) to provide order information.
- Accepts rich responses (buttons) from users.
Publications
How Do We Represent Data In Neural Networks
- Data representation for neural networks using NumPy arrays.
Democratizing Deep Learning
- For deep learning to reach full potential, we need to radically democratize it.
Deploying Rasa X On Linux Azure VM
- Deploying your Minimal Viable Assistant on Rasa X.