Selected work

Built around the problem.

A selection of analytics, automation, reporting, and data-product work. Client and employer names are intentionally omitted where the underlying work is confidential.

01REPORTING AUTOMATION

Enterprise compliance reporting automation

Rebuilt a recurring compliance-reporting workflow so completion metrics, manager rollups, operating-unit summaries, and branded outputs could be generated consistently from messy source data.

Challenge

The process depended on repetitive spreadsheet work across employee, manager, course, and organizational data, making monthly reporting slow and difficult to standardize.

Approach

Created automated data-cleaning, mapping, aggregation, and output workflows with business rules built directly into the process.

Result

Reduced manual reporting effort and created a repeatable reporting system that could scale across multiple business segments.

PythonPandasExcelVBAData Automation
02GEOSPATIAL ANALYTICS

Interactive map for 1,200+ operating locations

Turned a large operating-location dataset into an interactive geographic view with employee counts, management context, addresses, and location-level information.

Challenge

A spreadsheet could show where locations existed, but it could not make geographic patterns, concentration, or location context immediately visible.

Approach

Built a browser-based location intelligence tool using Leaflet and GeoJSON, with dynamically sized markers and contextual information available at each location.

Result

Made a nationwide operating footprint easier to explore and gave stakeholders a much faster way to understand location-level context.

JavaScriptLeafletGeoJSONPythonData Visualization
03PEOPLE ANALYTICS

Talent and succession analytics

Developed automated talent-mapping tools combining performance, potential, promotion readiness, retention risk, and impact-of-loss measures.

Challenge

Talent decisions required multiple dimensions of assessment data to be interpreted together while still producing an output leaders could understand quickly.

Approach

Built scoring logic, segmentation rules, and automated visual outputs that translated assessment inputs into structured talent views.

Result

Created a consistent framework for comparing talent information and turning qualitative assessment inputs into decision-ready outputs.

PythonExcelPeople AnalyticsScoring ModelsAutomation
04ASSESSMENT REPORTING

Automated 360° feedback reporting system

Built a reporting pipeline that converts multi-rater feedback data into individualized PDF reports with competency summaries and compact visualizations.

Challenge

Raw assessment results were useful analytically but too cumbersome to turn into polished participant-facing reports by hand.

Approach

Automated the data transformation, visual reporting, and PDF generation process using reusable templates.

Result

Made individualized reporting repeatable while preserving a consistent presentation across participants and competency areas.

PythonPandasJinja2WeasyPrintPDF Automation
05PSYCHOMETRICS

Multi-dimensional assessment analytics

Analyzed personality, career-interest, and item-level assessment data using multidimensional scoring, percentile interpretation, and item-response modeling.

Challenge

Assessment datasets contained multiple scales, item formats, and latent dimensions that could not be summarized meaningfully with a single score.

Approach

Combined statistical modeling and structured reporting to create interpretable profiles while preserving the multidimensional nature of the data.

Result

Produced richer assessment insights and a framework for translating complex psychometric results into understandable outputs.

RIRTPythonStatisticsAssessment Analytics
A note on confidentiality

Many analytics projects involve internal employee, operational, or business data. These examples focus on the problem solved, the technical approach, and the type of outcome rather than exposing proprietary datasets or confidential organizational details.

Have a similar problem?

If reporting takes too long, your data lives in too many places, or the information you need is hard to see, there is probably a cleaner way to build it.

Start a conversation →