Embedded Files
Turning data into decisions.
Turning data into decisions.
I specialize in uncovering insights that drive smarter strategies and measurable impact. My portfolio highlights projects across industries from financial services and e‑commerce to public policy and community programs where I’ve applied SQL, Python, Tableau, Looker Studio, and advanced statistical methods to solve real‑world problems.
BRICS Astronomy Data Analytics Capstone Project
BRICS_ASTRONOMY_DATA_ANALYTICS_CAPSTONE_PROJECT.pdfA project to analyze health data and identify stroke risk factors by organizing and modeling socio-demographic and clinical variables. The work involved cleaning raw medical inputs, validating features, and building predictive models supported by structured visualizations. This analysis produced clearer insights into stroke determinants, helping stakeholders anticipate risks, design prevention strategies, and strengthen public health decision-making. The project was completed as part of the BRICS Astronomy Data Analytics Capstone, which required demonstrating proficiency in data cleaning, exploratory analysis, modeling, and interpretation using real-world datasets.
Tools: Python, Pandas, NumPy, Scikit-learn, Matplotlib, SeabornAnalysis: Data Cleaning, Exploratory Data Analysis (EDA), Feature Engineering, Predictive Modeling, Visualization
Using the Kaggle Stroke Prediction Dataset, I applied the PACE framework to transform raw health records into actionable insights. I planned the analysis around identifying key stroke risk factors, cleaned and validated socio‑demographic and clinical variables, and conducted exploratory analysis to uncover trends. I then constructed predictive models with Python, Pandas, NumPy, and Scikit‑learn, and executed clear visualizations using Matplotlib and Seaborn. This approach highlighted the impact of age, hypertension, and lifestyle indicators on stroke occurrence, demonstrating advanced data science methods for public health decision‑making.
01 Project: Category Spreadsheet Analysis
CATEGORY SPREADSHEET FOR FINAL REPORT 3RD MILESTONE B00 TOWERS .xlsxA project to streamline milestone reporting for the B00 Towers initiative by organizing and analyzing categorical data. The work involved cleaning raw inputs, validating categories, and building structured dashboards to ensure clarity in project tracking. This analysis resulted in a more reliable reporting framework, helping stakeholders identify gaps, monitor progress, and allocate resources more effectively.
Tools: Microsoft Excel, Google Sheets, SQL, Python Analysis: Data Cleaning, EDA, Segmentation, Pivot Reporting
FOR DRONE SURVEY FINDINGS B01 185 To 175.docx.pdfThis analysis was based on raw drone survey data collected along the Matambo line. The dataset included tower coordinates, severity scores (1–5) for corrosion, metal unions, conductor bundles, arching horns, and insulator strings, plus field annotations.
The raw data was cleaned and standardized, then transformed into the Category Excel reporting framework. This process demonstrates my ability to handle the full data lifecycle from raw inspection records to structured dashboards and actionable insights.
02 Project: Bellabeat Case Study – Smart Wellness Insights
Bellabeat_case_study_2.docxA project to analyze Fitbit fitness, sleep, and weight datasets to uncover trends in smart device usage and provide actionable recommendations for Bellabeat’s marketing strategy. The work involved cleaning raw inputs, merging multiple datasets, and building structured visualizations to highlight user activity, sleep behavior, and calorie expenditure. This analysis revealed clear insights into consumer habits, helping Bellabeat identify opportunities to tailor product features and optimize marketing campaigns.
Case-Study-2_-How-can-a-wellness-technology-company-play-it-smart-.pdfData Source NoteThis case study was based on the FitBit Fitness Tracker dataset (CC0: Public Domain), containing daily activity, sleep, and weight logs from 30 users. The raw data was cleaned, merged, and transformed into the reporting framework above, demonstrating the full data lifecycle from raw CSVs → structured R analysis → actionable insights.
Case study 2 steps for analysis.docxIn this project, I followed the Ask–Prepare–Process–Analyze–Share–Act framework to guide the data analysis. I cleaned and transformed Fitbit activity, sleep, and weight datasets using SQL and R, ensuring accuracy and consistency.
Finally, I visualized key trends in Tableau and Excel to generate actionable insights for Bellabeat’s marketing strategy.
Finally, I visualized key trends in Tableau and Excel to generate actionable insights for Bellabeat’s marketing strategy.
Tools: RStudio (tidyverse, ggplot2, dplyr, lubridate, tidyr, purrr), SQL, Tableau, Excel
Analysis: Data Cleaning, Dataset Merging, Exploratory Data Analysis (EDA), Statistical Summaries, Visualization
Analysis: Data Cleaning, Dataset Merging, Exploratory Data Analysis (EDA), Statistical Summaries, Visualization
BellaBeat_Dashboard.pdfThis dashboard highlights key wellness metrics from Fitbit data, including active minutes, average weight, sleep duration, and BMI trends. It transforms raw activity and health logs into clear visual insights, making patterns in daily behavior easy to interpret. The visualization supports data‑driven recommendations for improving user wellness and guiding Bellabeat’s marketing strategy.
03 Project: Cyclistic Case Study – Bike‑Share Usage Insights
DA-C8-Case-Study-1-PDF.pdfA project to analyze historical trip data from Cyclistic, a Chicago bike‑share program, to uncover differences between casual riders and annual members. The work involved cleaning and transforming ride data, calculating ride length and weekday usage, and building pivot tables and visualizations to highlight behavioral patterns. This analysis revealed clear trends in rider behavior, supporting marketing strategies aimed at converting casual riders into annual members.
Tools: SQL, RStudio, Excel, Tableau Analysis: Data Cleaning, Ride Segmentation, Exploratory Data Analysis (EDA), Visualization, Business Recommendations
Case_study_Steps_For_Analysis.docxIn this project, I followed a structured workflow to analyze Cyclistic’s bike‑share data. I began by consolidating quarterly datasets, cleaning inconsistencies, and standardizing labels for members and casual riders. Additional fields such as ride length, date, and weekday were created to enable aggregation, and invalid records were removed.
This prepared dataset was then used for descriptive analysis and visualizations, revealing clear behavioral differences between casual riders and annual members to guide marketing strategy.
This prepared dataset was then used for descriptive analysis and visualizations, revealing clear behavioral differences between casual riders and annual members to guide marketing strategy.
04 Project: Digital Health Information Systems – Child Stunting Management in Malawi
IJISRT26JAN663.pdfA project to design and evaluate a digital health information system (DHIS) for managing child stunting interventions in Malawi. The study engaged 60 frontline stakeholders and demonstrated that a DHIS significantly improves data accuracy, reporting speed, and decision‑support compared to manual systems. Findings showed that digitization empowers social workers and health officers to deliver more timely, evidence‑based interventions, strengthening child nutrition outcomes.
Tools: Structured Questionnaires, Excel (Data Cleaning & Analysis), Descriptive Statistics Analysis: Quantitative Cross‑Sectional Study, Stakeholder Perception Analysis, Evidence‑Based Recommendations
05 Project: NYC TLC Data Readiness Project
Course-2-Automatidata-executive-summary.pptxThe presentation summarized the 2017 Yellow Cab dataset inspection (22,699 records). It highlighted critical issues such as negative fares (–$120), extreme outliers ($1,200 for 2.6 miles), and invalid passenger counts (zero passengers). Recommendations included converting timestamp fields to datetime objects, removing anomalies, and documenting all cleaning decisions to ensure reproducibility.
2017_Yellow_Taxi_Trip_Data (2).xlsxThe dataset contained trip details including pickup/dropoff times, trip distance, passenger counts, payment type, and fare components. While structurally complete with no null values, the data required cleaning before modeling. Key variables identified for predictive modeling were trip_distance (strong linear driver of fare) and payment_type (critical for tip prediction).
Activity_Course 2 Automatidata project lab.ipynb - Colab.pdfUsing Python (pandas, NumPy), the dataset was loaded into a dataframe, inspected with .info() and .describe(), and sorted to reveal anomalies. Outlier detection exposed unrealistic fares and zero‑distance trips with high totals. Analysis confirmed that credit card payments averaged $2.72 in tips vs. $0.00 for cash, making payment_type a key categorical feature. The code also grouped by vendor IDs, showing consistent average total amounts across providers.
Seoul Bicycle Rental – Tableau Visualization
Seoul-bicycle-rental-dataset.xlsxDataset Preparation
Using the Seoul Bicycle Rental dataset, I explored hourly rental records across weekdays in 2018, segmented by season. The dataset contained over 37,000 records, which I cleaned and organized to ensure consistency in time and seasonal variables. This preparation allowed me to highlight demand cycles and seasonal variations in bike usage.
Using the Seoul Bicycle Rental dataset, I explored hourly rental records across weekdays in 2018, segmented by season. The dataset contained over 37,000 records, which I cleaned and organized to ensure consistency in time and seasonal variables. This preparation allowed me to highlight demand cycles and seasonal variations in bike usage.
The dashboard demonstrates how Tableau can transform raw operational data into decision‑ready insights, helping city planners and transport managers optimize bike availability and anticipate seasonal demand.
Sheet 2.pdfI built an interactive dashboard in Tableau to visualize:
- Hourly Rental Trends – showing peak demand during morning and evening commute hours.
- Seasonal Variation – rentals highest in summer/autumn, lowest in winter.
- Weekday Patterns – consistent demand Monday–Friday, with slight weekend deviations.
06 Internship Project: Telecom Customer Churn Analysis
A project to analyze customer churn in a telecommunications company by applying structured business analysis tasks. The work involved loading and cleaning raw customer data, validating records, and preparing the dataset for deeper insights. Exploratory analysis highlighted churn proportions, while segmentation visualizations revealed tenure‑based customer patterns. Advanced analysis examined churn by demographics, payment methods, and contract types, resulting in actionable recommendations to improve retention.
The raw dataset contained customer demographics, service usage, contract details, and churn status. This dataset provided the foundation for all analysis and modeling, showcasing the ability to work with real‑world structured data and prepare it for analytics.
The raw dataset contained customer demographics, service usage, contract details, and churn status. This dataset provided the foundation for all analysis and modeling, showcasing the ability to work with real‑world structured data and prepare it for analytics.
Tools: CSV Data, Python, Pandas, Matplotlib, Seaborn, Plotly Analysis: Data Loading, Preprocessing, Feature Exploration, EDA, Segmentation, Statistical Modeling, Demographic Analysis
BUSINESS ANALYSIS TASK LIST.pdfA structured task list provided by SaiKet Systems to guide the internship project. It outlined five key stages: dataset understanding, data cleaning, exploratory analysis, customer segmentation visualization, and advanced churn analysis. This framework ensured clarity of scope and helped align technical work with business objectives.
Tools: Documentation, Task Planning Analysis: Requirement Structuring, Workflow Definition
Saiket_systems_Data_analytics.ipynb - Colab.pdfA Jupyter/Colab notebook implementing the full churn analysis pipeline. The notebook demonstrates proficiency in Python data handling, visualization, and statistical analysis. It includes customer segmentation pie charts, clustered bar charts of monthly charges, and churn rate comparisons across demographics and contract types.
Tools: Python, Pandas, Seaborn, Matplotlib, Plotly Analysis: Data Preparation, Visualization, Comparative Analysis
Page updated
Google Sites
Report abuse