Case Studies

Projects & Results

A selection of the types of engagements I take on. These represent realistic project scopes, outcomes, and approaches — updated with real client work as engagements are completed.

01
Python AutomationProfessional Services

Automated Financial Reporting Pipeline

A professional services firm was spending four hours each month manually compiling a financial performance report from three separate data sources. The process was error-prone, time-consuming, and always happened under deadline pressure.

4 hours
reduced to under 5 minutes
Zero errors
in post-launch reports
Monthly delivery
fully automated

The Challenge

The client had data spread across an accounting platform, a CRM, and a project tracking tool. Reconciling these sources required manual exports, VLOOKUP-heavy spreadsheets, and significant staff time every month. The final report was often late and occasionally contained errors that required rework.

The Solution

I built a Python pipeline that authenticates with each data source via API or direct database connection, pulls the required data on a schedule, applies the reconciliation logic, and outputs a formatted Excel report delivered automatically to the relevant team members.

Built with:PythonpandasopenpyxlREST APIsSQLAlchemyCelery
02
Predictive AnalyticsRetail

Sales Forecasting Dashboard for Retail SMB

A retail business was making purchasing decisions based on prior-year sales and intuition. Without a systematic forecasting approach, they routinely faced overstock on slow-moving products and stockouts on high-demand items.

23%
reduction in overstock costs
90-day
rolling forecast window
Weekly refresh
automated data pipeline

The Challenge

The business had three years of transaction-level sales data but no systematic way to use it for planning. Seasonal patterns, promotional effects, and supplier lead times were all tracked informally. Purchasing decisions were delayed and often reactive.

The Solution

I built a demand forecasting model using historical sales data, seasonal decomposition, and promotional event flags. The model outputs 90-day rolling forecasts at the product-category level, surfaced in a Streamlit dashboard the operations team could access directly.

Built with:Pythonscikit-learnpandasStreamlitPlotlyPostgreSQL
03
AI Web ApplicationLogistics & Operations

Internal Operations Dashboard

A growing operations team was managing their day-to-day workflow across three disconnected tools: a project tracker, a CRM, and a custom spreadsheet. The result was constant tab-switching, manual data transfers, and a management team with no reliable real-time view of operations.

3 systems
unified into one view
Real-time
operational visibility
2-week
build and deployment

The Challenge

The client needed a single operational view that unified data from their existing systems without replacing them. The solution had to be fast to build, easy to maintain, and usable without technical training. Off-the-shelf BI tools were evaluated but could not handle the custom data relationships involved.

The Solution

I built a lightweight internal web application with a FastAPI backend and a React frontend. The app pulls and normalizes data from all three systems on a refresh cycle, presents a unified operational dashboard, and includes basic filtering and drill-down by team, project, and date range.

Built with:PythonFastAPIReactNext.jsPostgreSQLREST APIs

Sample case studies — updated with real client work as engagements are completed.

Start Your Project

Have a Similar Problem?

Every engagement starts with understanding the specific problem. Reach out and we will talk through whether it is something I can help with.