From Data to Insights

This page documents projects I've built in the real world — production pipelines, automation systems, and analytical tools deployed for actual users. It also includes earlier certification work that shaped the foundation. The common thread is the same: take a messy problem, understand what the data is actually saying, and build something that makes the answer usable. 

Power BI Export Cleanup & Email Automation:

One process at Dynatrace was manually pulling and reconciling data from four separate Power BI exports to track account risk — a slow, error-prone process with no consistent format across teams. I built an automated Python pipeline that merges all four sources into a master dataset, scores accounts by renewal risk, and generates personalized filtered Excel files for each AE, CSM, and Renewal Rep with only their relevant accounts included.

To distribute those files, I built VBA macros that automate workbook setup and formatting, then ran a mail merge pipeline that attaches the right Excel file to the right person's email automatically eliminating manual sorting and sending entirely.

I'm currently extending the system with an AI layer for an internal contest (due May 2025) that reads each account's data and generates a plain-English summary with recommended next steps making the insights actionable for reps who don't live in spreadsheets.

Physician Scheduling Automation:

The scheduling team at IHA was spending over two months manually building 6-month physician schedules across five hospitals — cross-referencing FTE contracts, provider preferences, PTO, credentialing rules, and role eligibility by hand. I replaced that process with a Python-based scheduling engine that handles it automatically.

The system ingests structured data tables from Excel, applies a multi-pass constraint-satisfaction algorithm, and outputs a production-ready CSV for direct import into Momentum (their scheduling platform). It enforces 550+ normalized provider preferences, weighted night-shift balancing, rest buffer rules, monthly distribution leveling, and triage role constraints — then generates diagnostic reports so any gaps are immediately visible.

I also built a macro-enabled Excel calendar view and availability tracker for management review, and designed a Microsoft Teams intake form to normalize provider preference requests going forward.

Previous Build Approach:

The team copied and pasted all of the providers' preferences that were in paragraph format in Momentum into an Excel Workbook, and referenced the preferences when individually and manually inputting each provider into Momentum.

My Build:

Below: the normalized constraint data model in Excel that fed the scheduling engine, and the Python script that processed it — building a complete 6-month schedule in minutes. (All data shown is HIPAA compliant — provider identities are publicly listed medical staff.) 

During my journey to earning Microsoft certifications in Power BI, SQL, and Excel, I worked on various projects that strengthened my ability to transform raw data into meaningful insights. Through hands-on experience, I developed skills in data modeling, visualization, and query optimization, applying them to real-world scenarios. 

Retirement Financial Planner:

Below is a custom Excel-based financial planning tool I created for a client who wanted a clear, interactive way to track their income, expenses, and assets while calculating the amount they’d need to withdraw from investments each month to break-even.

While similar tools exist through financial advisors, my client wanted full control—to see the calculations firsthand and update their data as life circumstances changed. To achieve this, I designed a dynamic Excel template using placeholder names (“John” and “Jane”), where all sheets are interconnected through formulas. The dashboard provides a clear summary, including projections up to their estimated life expectancy.

To enhance the insights, I then imported the data into Power BI, where I created custom measures and visualizations to help the client better understand their financial outlook at a glance.

Bicycle Company Revenue Analysis:

As part of my Power BI Microsoft Certification, I worked with a dataset from a bicycle company, where I had the flexibility to create custom measures, calculations, and visualizations. 

One transformation I implemented was a query that adjusted revenue based on each country's currency, ensuring accurate financial analysis across different markets. To present the insights effectively, I designed a dynamic dashboard featuring: A column chart to visualize the net revenue trends, a line chart for tracking gross revenue performance over time, a data table with essential metrics, and slicers for filtering by country and date, allowing for interactive exploration.

This project strengthened my ability to transform raw data into meaningful insights using Power BI.

Beyond these projects, I’ve applied SQL for data extraction and manipulation and have been learning Python to further automate and enhance my analytical capabilities. My experience across these tools has equipped me to tackle complex data challenges and deliver actionable insights.