About Me

A dedicated Data Analyst with 2 years and 8 months of robust experience at Hong Kong MEXIMCO Group Limited, I excel in extracting pivotal business insights from intricate datasets, fostering growth and refining operational efficacy. Successfully employing statistical models, I've forecasted sales trends with a notable 72% precision and spearheaded projects culminating in a 24% surge in trading transactions and 21% revenue enhancement. Having graduated with a Master of Science in Business Analytics from HKU Business School, I integrate my engineering acumen with proficiencies in Python, R, and SQL to craft strategic data resolutions. Fluent in AGILE, Big Data Analytics, and Machine Learning paradigms, I've crafted web apps and dashboards that accentuate data visualization and facilitate informed decisions. Backed by a repository of industry certifications from leading platforms like LinkedIn, freeCodeCamp, and HarvardX, I'm poised to address sophisticated data conundrums. An adept communicator, my expertise spans liaising with diverse teams, engaging stakeholders, and unveiling actionable data narratives. Bilingual in English and Bengali, I am primed to drive data-centric innovations in multifaceted settings.

Education

HKU Business School, The University of Hong Kong (HKU)

August 2022 - August 2023

Master of Science in Business Analytics, MSc(BA)

Entrance Scholarship, Student Representative of the 2022-23 MSc(BA) batch

Electives Taken: Machine Learning, Database Management System (DBMS), Social Media and Marketing Analytics, Deep Learning, Big Data Analytics on the Cloud

Professional Experience

Data Analyst (Remote)

Hong Kong MEXIMCO Group Limited

September 2019 - February 2020

Established advanced statistical models to identify customer trends and predict future sales growth with 72% accuracy.
Conducted statistical analysis and built predictive models to uncover trends in export transaction data, resulting in a 24% increase in more incoming trading.
Constructed and implemented data quality checks that reduced errors in reports by 22%.
Carried out ad-hoc analysis to answer business questions and provided insights to stakeholders, resulting in improved operational efficiency and cost savings.
Collaborated with cross-functional teams to analyze large volumes of global export trading data to uncover key insights, leading to a 21% increase in revenue for the company.

Data Analyst Intern (Remote)

Hong Kong MEXIMCO Group Limited

March 2020 - April 2022

Applied Data Visualisation Tools such Tableau and Orange Data Mining to create interactive graphical representations of financial and trading data.
Constructed Data Models and documented data flows to improve system architecture and data quality.
Collaborated with cross-functional teams to identify and resolve data-related issues, resulting in a 16% improvement in data accuracy.

Featured Projects

Customer Segmentation

Identified distinct customer segments, created four customer personas, and developed more effective marketing strategies

View Project

NLP Project: Native/Non-Native Speaker Classifier

This NLP project analyses unstructured text data from both Native and Non-Native Speakers and attempts to build a Binary Classification Model with the Textual features

View Project

Employee Attrition Classifier

Identified factors that contribute to employee attrition, analysed employee data to determine patterns or trends, and developed strategies for reducing employee turnover and improving retention.

View Project

Wine Quality Testing

Machine Learning Techniques have been applied to predict the Wine Quality level. A collection of Regression and Classification Models have been used. The XGBoost Regressor Model has been found to be the best model. The XGBoost Regressor Model has been hyperparameter tuned. The RMSE value of the XGBoost Regressor Model after tuning did not show any prominent impact. The Dataset was subjected to Feature selection before analysis

View Project

Web-Scraping Tabular Data

A webpage containing tabular data is webscraped and the information table is outputted as a CSV file.

View Project

Daily Stock Price Dashboard

A dashboard to view the stock price of Apple Inc.

View Project

Netflix Dashboard

A dashboard that provides insights into Netflix titles, including ratings, genre, release date, and popularity. This dashboard can be used to help users find new titles to watch, track their viewing habits, and learn more about the Netflix library.

View Dashboard

Simple Classifier Exploration

A web app made with python modules numpy, scikit-learn, matplotlib, and streamlit to allow exploration of different classification models on different datasets. The model has been deployed as a webapp on streamlit.

View Project

Daily Stock Price Dashboard (Web App)

A web app made with python modules pandas, yfinance, Ploty, and streamlit to allow viewing of the stock prices of Apple, Google or Microsoft. The model has been deployed as a webapp on streamlit.

View Project

Loan Data Analysis with KNIME

Machine Learning Techniques have been used to investigate the prediction of loans being full paid or not. The KNIME analytics platform has been used to demonstrate the utilisation of visual programming in achieving this task. A Random Forest Classifier has been used and the accuracy for this model is 84.2%. A Decision Tree Classifier has been used and the accuracy for this model is 84.3%. A Naive Bayes Classifier has been used and the accuracy for this model is 84.2%. All three applied models show the nearly identical accuracy.

View Project

View Dataset Used

Data Science Salaries Analysis with SQL

Explored the Data Science Salaries Dataset with SQL

View Project

Titanic Survivor Analysis

The Titanic survivor analysis is a machine learning projects that uses data from the titanic disaster to predict where a passenger survived. The project features a variety of features such a passenger's age, gender, class and port of embarkation, to train a model that can predict survival with an accuracy of over 80%.

View Project

Data Science Salaries Dashboard

The dashboard visualizes data from the Kaggle dataset on data scientist salaries. It uses charts and graphs to show how salaries vary by job title, company, location, and experience level. The dashboard also includes a table of data that allows users to filter the data by specific criteria. The dashboard is interactive, so users can click on the charts and graphs to see more detailed information. Overall, the dashboard is a useful tool for understanding the data science salaries landscape.

View Project

Top 50 Fast Food Franchises (USA) Dashboard

A Dashboard which draws insights from the dataset and illustrates how the top 50 fast food franchises in the US have been faring in the year 2021 from the year 2020. Skills: Exploratory Data Analysis, Tableau, Data Visualisation

View Dashboard

Facebook Marketing Campaign Dashboard

A Streamlit Web App of a Facebook Marketing Campaign Dashboard. Skills: Data Visualisation, Python3, Plotly, Streamlit

View Web App

View Project

PyGWalker Demo Web App on Streamlit

A web app to demo PyGWalker with an uploaded CSV file

View Web App

View Project

Developer Salary Prediction ML Web App

A web app built with Streamlit and Python programming which allows the exploration of a dataset and using said dataset to make predictions on its target column.

View Web App

View Project

Analysis of NYC Public School Test Scores

PostgreSQL has been used to explore the NYC Public School Test Scores Dataset

View Project

Sample Employee Database SQL Exploration

MySQL/SQL has been used to explore and draw insights from the Sample Employee Database.

View Project

Finance Stock Dashboard

A streamlit web dashboard app to inspect stocks and see their latest news.

View Project View Web App

A/B Testing (in Python)

A/B Testing Experiment done in Python.

View Project

Multi-Class Classification

A Multi-Class Classification on the UCC Unhealthy Comments Dataset implemented in Python's PyTorch and HuggingFace.

View Project

Countrywise Production-based CO2 Emissions Dashboard

A Tableau Dashboard showing the Global CO2 Emissions from various countries across the globe.

View Dashboard

Credit Card Approval Classification

This is a machine learning project that analyzes the Kaggle Dataset and does a classification task of determining whether a credit card has been approved or not.

Technology used: Python, Orange Data Mining, Version Control, Machine Learning, Classification Algorithms

View Project

Data Professional Survey Dashboard Report

A Dashboard (made with Power BI Desktop) crafted from real data of Data Professionals during the year 2022.

Skills Used: Data Preprocessing, Data Wrangling, Data Visualization

Technology used: Power BI Desktop

View Project

Space Missions Dashboard Project

A Dashboard made with Power BI Desktop based on the Space Missions Dataset from Maven Analytics.

Skills Used: Data Visualization

Technology used: Power BI Desktop

View Project

Relevant Certifications

Project Management (Specialization)

Issuing Organisation: Univerity of Colorado Boulder

Platform: Coursera

Date Acquired: November 2023

View Certification

Prompt Engineering for ChatGPT

Issuing Organisation: Vanderbilt University

Platform: Coursera

Date Acquired: October 2023

View Certification