What I Bring to the Table
I design data systems that drive action and unlock impact. My foundation in research psychology and statistics gives me the tools to uncover patterns in messy, real-world data. Whether in education, nonprofit work, or operational analytics environments, I specialize in building models, dashboards, and AI-driven tools that turn complexity into clarity.
I bring a blend of experimentation, analytical rigor, and the ability to see problems from multiple vantage points. Whether it's uncovering risk patterns in complex populations, aligning resources with forecasted demand, or building agentic AI tools that turn ambiguity into action, I approach each challenge with both technical precision and strategic awareness. My work is adaptable, outcome-oriented, and always focused on enabling better decisions, regardless of the domain.
Where I Excel
Predictive Modeling & Feature Engineering
Forecasting & Time Series Modeling
A/B Testing & Experimental Design
Agentic AI & LLM Workflow Design
Root Cause & KPI Change Analysis
MLOps & Model Monitoring (MLflow, Evidently)
SQL, Python, Databricks
Dashboard Design (Tableau, Power BI)
Stakeholder Communication & Enablement
Selected Impact Highlights
Predictive Modeling & AI
Reduced marketing spend by an estimated 5% by deploying a daily enrollment propensity model that identified high-likelihood applicants, enabling better outreach decisions and resource targeting.
Enabled future budget planning with a 16 month forecast built using a hybrid XGBoost pipeline, informing programming scenarios.
Accelerated project intake and scoping through a personally built Agentic AI chatbot that converts free-text requests into structured project charters with risk scenarios and assumptions.
View Project Estimated long term ROI using a forecasting model trained on only 5 months of production data, with just 14% error (nMAE), demonstrating strong predictive power even with limited history.
View Project
Enabled enrollment to better assess the impact of its business processes by building an enrollment model to determine the applicants natural propensity to enroll at the time of the application.
Optimized the new start forecast by leading the MLOps migration of 50+ notebooks into the MARVIN MLOps framework.
Dashboards & Decision Support
Prevented a 10% enrollment loss by creating a real-time Daily Enrollment dashboard that detected a mid-cycle dip. The alert prompted early marketing outreach, reducing the projected summer decline from 10% to 2%.
Reduced funnel attrition by 3% with a Tableau dashboard that surfaced key drop-off points in a new self-service enrollment process using control charts and KPI alerts.
Improved applicant engagement timing by running A/B tests and visualizing the results in a Tableau dashboard, revealing time-sensitive effects that helped marketing teams refine outreach strategies for better impact.
Improved solar system performance monitoring by developing a Tableau dashboard that tracks solar panel production over time, enabling early identification of performance drops and giving stakeholders an intuitive view of system effectiveness.
View Project
Improved executive decision-making by designing 40+ Power BI dashboards across HR, finance, and academics at McHenry County College, used by the President's cabinet and board.
Increased transcript brokerage program adoption by creating a dashboard that showed students using a transcript brokerage service enrolled more quickly than the comparison group by 56% prompting broader promotion and support from leadership.
Leadership & Enablement
Advised the college president directly on high-stakes strategy and policy questions using live dashboards and real-time data storytelling at McHenry County College.
Translated data to action for the board by presenting institutional performance and program ROI results in plain language, influencing budget and policy.
Led cross-functional agentic AI design for a chatbot that transforms vague project requests into actionable charters, guiding data science scoping at scale within WGU.
Improved team capability and autonomy by mentoring 4 analysts in SQL, Tableau, and experimental design, resulting in stronger dashboards and fewer ad hoc requests.
Boosted operational awareness by creating dashboards with visual triggers (like control chart alerts) and supporting them with targeted walkthroughs for enrollment teams, increasing confidence in data use and responsiveness.
Brought clarity to ambiguous executive requests by turning vague, high-stakes questions from leadership into structured analysis and decision-ready dashboards, ensuring fast alignment without needing repeated clarification.
Track Record
- Western Governors University – Data Scientist & Analytics Manager (2022–Present)
- McHenry County College – Director of Institutional Research (2012–2022)
Agentic AI in Action
The Project Charter Assistant is an intelligent intake tool I designed and built using LLM agents. It’s available on my Critical Influence site and is intended to support both analysts and non-technical stakeholders. The assistant interprets free-text project requests, fills in gaps with best-practice prompts, and auto-generates full business-aligned project charters with scenario plans, risks, and success criteria, helping teams clarify scope, expectations, and assumptions before data work begins.
This app demonstrates my ability to apply language models with structured architecture and UX awareness. I designed the JSON intake flow, created prompt logic, implemented API calls, and built the front-end experience — all hosted in Firebase.
Try the Charter AssistantWhat Others Say
From My Mentor
“I’ve had the pleasure of mentoring Amy Humke, and her growth has been nothing short of extraordinary. She quickly established herself as a capable and driven data scientist, building and deploying a complex classification model that supports Enrollment strategy. Through that project, she deepened her skills in exploratory data analysis, feature engineering, and advanced modeling techniques—including model stacking, data balancing, and hyperparameter tuning. She later cloned and adapted the model for a different use case in Marketing, a testament to its robustness and versatility. She also became the first on our team to migrate a model to our MLOps platform and has since worked with nuanced modeling techniques like soft targets and addressed challenges like data drift. More recently, she’s been expanding into LLMs and has taken a leadership role in our chatbot project, a cross-functional effort shared by four data scientists. What sets Amy apart is her ability to learn fast, execute with precision, and lead with clarity. Her contributions have already been recognized at the highest levels, and any team would be lucky to have her.”
— Alberto Evangelista, Senior Data Scientist
From My Manager
“Amy Humke is an outstanding data scientist and analytics leader who has made significant contributions to Western Governors University (WGU) through her developing data science expertise and dedication to the institution's mission. Most notably, Amy spearheaded the development of an Enrollment Propensity Scoring model, a complex project that involved the application of cutting-edge machine learning techniques. Her efforts not only improved the model’s F1 score from 33% to 48.5%, but also laid a strong foundation for predictive analytics at WGU, enabling leadership to better understand and optimize enrollment processes. Additionally, Amy played a pivotal role in the creation of the Enrollment Regionalization Dashboard, which has been instrumental in aligning staffing resources with call volume trends, thereby optimizing enrollment counselor support. Amy’s ability to overcome significant challenges, such as navigating new ML Operations platforms and managing complex data pipelines, demonstrates her exceptional data analytical skills, innovative mindset, and commitment to continuous improvement. Her leadership in mentoring team members and fostering a collaborative environment further underscores her role as a key contributor to WGU’s success. Amy would be an invaluable asset to any team, bringing a unique blend of technical expertise, detail-oriented thinking, and leadership.”
— Eric Hansen, Director of Analytics, ex-Adobe
From a Colleague
“Amy's exceptional skills as an Enrollment Analytics Manager have been evident in every project we've collaborated on. Her expertise in SQL, combined with her strong data science aptitude and advanced Tableau skills, consistently produces insightful analyses and compelling visualizations. Amy’s creativity and willingness to assist others make her a valuable asset to any team, driving innovation and success. In addition to her technical expertise, Amy has played a crucial role in my professional development. Her guidance has significantly enhanced my skills in Tableau, SQL, and Python, which has been pivotal in advancing my data science career. Amy's dedication to mentoring and her ability to foster growth in others speak to her strengths as both a leader and a colleague.”
— Rinky Devre, PhD, Research Analyst
Ready to Work Together?
Looking for a data scientist who can deliver results and explain them? Let’s connect.
Email: amyhumke@gmail.com
LinkedIn: linkedin.com/in/amyhumke