Tunde Smith Amupitan

First-Class Honours Data Analytics graduate from Dublin Business School with hands-on experience in Python, SQL, Power BI, Tableau, and machine learning and current AWS Data Centre Security Vendor with a strong foundation in Machine Learning pipelines (Python, TensorFlow, Scikit-Learn). Skilled in the full analytics workflow, data collection, cleaning, analysis, visualisation, and deployment. Built and deployed a live sentiment analysis application using Streamlit, developed predictive models using Python, and created interactive sales and data visualisation dashboards. Strong foundation in statistics, data mining, and business intelligence gained through academic projects and real-world application. Proven capability in structured problem-solving, risk assessment, and managing complex workflows within high-security cloud environments. Adept at translating data insights into optimized operational processes, with deep academic and practical curiosity for scaling Generative AI, agentic workflows, and responsible automation governance.

[ 02/02/2026 – 05/04/2026 ]  Nollywood Sentiment Analysis | NLP & Streamlit Dashboard  Built a sentiment analysis model on Nollywood movie reviews using Python, NLTK, and Scikit-learn. Deployed an interactive dashboard using Streamlit that allows users to input text and receive real-time sentiment predictions. Demonstrates end-to-end data science Workflow: data collection → preprocessing → modelling → deployment. 

Link:  https://sentimental-analysis-krse8gdazsktpeyvhcpx6r.streamlit.app/?

[ 10/09/2025 – 30/12/2025 ]  Data Visualisation Project | Python & Tableau  Developed a series of data visualisations to explore and communicate trends within a real-world dataset. Used Python (Pandas, Matplotlib, Seaborn) for exploratory data analysis and Tableau for interactive dashboard creation. Project focused on translating raw data into clear, actionable insights for non-technical audiences

[ 02/03/2026 – 25/04/2026 ]  Energy Consumption Prediction Using Machine Learning  This project develops a machine learning workflow to predict electricity consumption using environmental and temporal variables. It applies K-Means clustering and polynomial regression (OLS, Lasso, Ridge) to both a synthetic toy dataset and real German electricity data from 2019, sourced from Open Power System Data and Open-Meteo. Lasso achieved the best performance with R²=0.9665, demonstrating that regularisation is essential for accurate and interpretable energy forecasting. 

Link:  https://colab.research.google.com/drive/1JP0i4zejWhpNT-3tqrLmcmpWa6xQbHMP? usp=sharing

[ 16/12/2025 – 10/02/2026 ]  Bankruptcy Prediction Using Machine Learning  This project develops a machine learning-based framework to predict corporate bankruptcy using financial ratios derived from publicly listed U.S. companies. Logistic Regression is applied on key indicators, including Altman Z-Score, ROA, Debt-to-Assets, and Working Capital ratios, across a balanced dataset of 37 bankrupt and 74 healthy firms. The model is rigorously evaluated through ROC-AUC analysis, 5-fold cross validation, and robustness testing against alternative model specifications.

 Link:  https://colab.research.google.com/drive/155u1f2vrYm_rYG0On-okjAi8qirBL-pP? usp=sharing

• Gender: Male
• Nationality: Nigerian
• VISA TYPE: STAMP 4

• AWS Certified Machine Learning - Specialty (MLS-C01) Cert Prep (https://www.linkedin.com/learning/certificates/6d6ffa1db9a2a4b2815eeca78225eab717a3ef2d0b04d6c4256cb3babce1c21f)
• Hands-On Data Analysis with ChatGPT-5 (https://www.linkedin.com/learning/certificates/94b7a4bfc9cb02c280f2eb61d725c3d57658013f81943f9d8c4a72e0b7a26e9b)

• https://github.com/Natipuma
• https://sentimental-analysis-krse8gdazsktpeyvhcpx6r.streamlit.app/
• https://colab.research.google.com/drive/1JP0i4zejWhpNT-3tqrLmcmpWa6xQbHMP?usp=sharing