Product demand for supply chain optimization - GenAi

Developed a Generative AI model to predict product demand, optimizing supply chain management for retail and manufacturing industries. The solution improved forecast accuracy and reduced inventory costs, addressing variability in demand patterns and enabling more robust decision-making under dynamic market conditions. 

Smart Ad A/B Testing

A/B testing is a user experience research methodology. A/B tests consist of a randomized experiment with two variants, A and B, which are identical except for one variation that might affect a user's behaviour. It includes the application of statistical hypothesis testing or "two-sample hypothesis testing" as used in the field of statistics.

Tech Stack: Inferential statistics, abtest

data science projects portfolio

This portfolio project is a combination of data science projects including Customer Churn, Data Collection, Data visualization, Image-text-data detection and Image filtering.  These Analyses are based on Data science basic technology. 

Technology: Selenium, statistics, matplotlib, pytorch

Twitter sentimental Analysis

Sentimental analysis helps us determine what people are talking about in a given area. Twitter helps us determine this by collecting data from a given area. We determine in this project by scoring the polarity, sensitivity and sentiments of covid-19 data and what is trending with covid 19. After this, I created a dashboard to help in visualizing this data with a word cloud and streamlit.

Tech Stack: streamlit dashboard, inferential statistics, Mysql

data Warehouse

The data uses data provided in https://anson.ucdavis.edu/~clarkf/ to create an ELT pipeline where transformations occur within the data warehouse through models written in the data build tool. The datasets are extracted from the links in the website and loaded into a PySpark data frame. The PySpark data frame is loaded into MySQL using JDBC. The extraction and loading have been done using Jupyter notebook. The models written in DBT are written using SQL syntax and Jinja. More transformations can be easily done on the data by writing a script for the required transformation. The transformation is done under the sensor_dbt_project folder.

Tech Stack: PySpark, MySQL, Airflow and DBT

Data science Micro degree

This Micro degree project focuses on python for Data Science and how to use Matplotlib, Numpy, Pandas and Seaborn.

Data Visualization using the tools for visualization, how to understand data and create plots and graphs

Date application

This application was one of my projects in React for dating. It looks like Tinder, and some of the features are like getting a hook-up, getting someone you like and connecting with them when they have an account with the platform. 

Tech Stack: React

Computer vision

In the booming industry of computer vision, in this project, I perform image segmentation and detection, feature calculation and object tracking. 

Telecommunication Data analysis

Performing a Data Overview, outlier detection, Data Preprocessing, cleaning data, Univariate analysis, Multivariate analysis, Correlation analysis, and bivariate analysis. Performance of the user engagement analysis, prediction and overview analysis helped determine the worth of a telecommunication company. 

Speech Data collection

Speech-to-text data transformation is important to understand NLP. In this project, I performed tasks of loading audio files and the transcriptions, converting them to channels, standardising and resizing them, doing data augmentation, feature extraction and acoustic modelling to create a Machine Learning model and a pipeline that can be used to collect and store the audio files and finally transform them to text data. 

Vehicle fatality

Mapping data with geopandas to identify where there is more accidents in a given area helps in reducing and establishing accident causes. In this project, I create a map and establish what are the causes of accidents in a given area and how to reduce them by making a causality inference to establish the causes and avoid them.  

Dashboards 

Creating Dashboards for machine learning and Data Engineering with streamlit and MLFlow is a great skill. Here are a number of dashboards that I have created with MlFlow and Streamlit to visualize data from different sources. In these, I show the skills for visualizing data and how to plot data in different ways.  I also show hoe to visualize machine learning models to show the way oin which they erfoem.