Portfolio / 2026

Bridging the gap between raw data chaos and actionable business strategy

Bridgette Heiner is a data scientist and systems thinker who uses statistical experimentation and predictive pipelines to solve complex operational challenges.

Open to thoughtful teams, data science roles, and conversations about what's next.

Editorial workspace with creative tools — Bridgette HeinerPortfolio preview

Cross-functional partner

Strategy to delivery

People-first systems

Selected Work & Technical Case Studies

How I turn complex problems into measurable progress.

Explore technical implementations, machine learning pipelines, and system architectures below.

Computer Vision · Deep Learning | Python · TensorFlow

Automating Waste Sorting with Image Recognition

Built an image classification model to help consumers correctly separate compost from recyclables and reduce landfill contamination. Benchmarked custom neural networks against pre-trained vision models to optimize accuracy.

Outcome / Achieved 91.6% sorting accuracy across 25,000+ images, validating computer vision for automated waste processing.

Generative AI · System Evaluation | NLP · LLMs

Improving AI Accuracy for Corporate Financial Intelligence

Evaluated an enterprise AI search system (RAG) designed to extract strategic financial insights across 160+ SEC filings. Diagnosed why the model gave incomplete answers and designed fixes to stop hallucinations.

Outcome / Delivered an actionable blueprint for document indexing and smart search, ensuring reliable data for executive decision-making.

Text Analytics · Predictive Modeling | PySpark · Machine Learning

Uncovering User Pain Points from 105,000+ App Reviews

Built a big-data text mining pipeline to analyze mobile reviews and uncover what drives customer frustration. Trained a predictive model to evaluate how support teams prioritize and answer critical complaints.

Outcome / Identified key drivers of churn (like billing friction) and built an 83% accurate model to predict support triage priority.

Workforce Analytics · Forecasting | R · Time Series

Predicting Employee Turnover and Headcount Demand

Analyzed enterprise HR data to identify why employees leave and uncover retention risk patterns. Modeled multi-year staffing trends to help leadership plan future hiring needs.

Outcome / Delivered a 76.5% accurate retention model paired with a 24-month headcount forecast to prevent staffing shortages.

A note for hiring teams

Looking for someone who turns complex data into measurable impact?

From distributed NLP and computer vision to predictive forecasting, I thrive on untangling messy data to build high-accuracy models.

Let's connect