Projects

Each of these projects tells a story—from solving business pain points to exploring AI’s creative frontiers. I’ve included a mix of industry work, academic research, and personal initiatives that highlight my strengths in machine learning, data engineering, and AI-driven automation.

Smart Manufacturing: Predictive Maintenance for Toyota

Role: Lead Data Scientist
Company: Centurylink (Lumen)
Tech Stack: Python, TensorFlow, Power BI, MLflow

I built predictive models to identify quality drops in Toyota’s injection molding process. Using historical sensor and defect data, I developed an ML pipeline to predict when product quality might decline, triggering early interventions. The models helped reduce downtime and improve first-time yield by 18%.

➡️ Impact: Boosted operational efficiency and reduced scrap costs.

Snowflake Data Warehouse Automation for Caterpillar

Role: Data Engineer
Company: Caterpillar
Tech Stack: Snowflake, Azure Data Factory, SQL, dbt

I automated ETL pipelines and built a scalable Snowflake data warehouse for Caterpillar’s drivetrain analytics. By implementing CI/CD practices and optimizing queries, I reduced data latency by 70% and operational costs by 20%.

➡️ Impact: Enabled real-time insights for global fleet analytics teams.

Foot Traffic Analysis using IoT and Video Data

Role: Independent Researcher
Tech Stack: Python, OpenCV, Deep Learning, Raspberry Pi, Power BI

This personal project involved setting up a camera-based system to count and classify foot traffic in public spaces. I used object detection models to track movement patterns and built a dashboard to visualize peak hours and heatmaps.

➡️ Impact: Potential applications in smart cities, retail optimization, and safety.

Wireframe to Website: Neural Network Translator

Role: Team Lead
Company: Centurylink
Tech Stack: PyTorch, CNNs, OCR

I designed and led the development of a deep learning model that translates hand-drawn UI wireframes into HTML web pages. This project explored the integration of CNNs with OCR and layout recognition to automate frontend design.

➡️ Impact: Time-saving tool for designers and developers.

Stock Market Chatbot for Retail Investors

Role: Developer & Data Scientist
Tech Stack: Python, NLTK, yFinance API, Flask

Created a chatbot that responds to user queries about stock performance, financial metrics, and trend analysis. The bot uses natural language processing to interpret questions and deliver insights using real-time financial data.

➡️ Impact: Made financial data more accessible to non-technical users.

Human Capital Analytics Dashboards

Role: Data Scientist
Company: HCMI
Tech Stack: Power BI, Azure Synapse, PySpark

Built dynamic dashboards to analyze workforce trends, attrition risk, and KPI forecasts. Implemented statistical models to predict future HR metrics and designed data models that connected multiple internal systems.

➡️ Impact: Supported leadership in strategic decision-making.

Data Anonymization for Privacy Preservation

Role: Research Assistant
Institution: University of Maryland, Baltimore County
Tech Stack: Python, ARX, K-anonymity, L-diversity

Published a comparative study evaluating anonymization techniques on large datasets. Demonstrated how to achieve 90% privacy retention while maintaining analytical value for research purposes.

➡️ Impact: Contributed to ethical data practices in academic and medical data sharing.