Hi, my name is
Nashville-based analytics professional operating at the crossroads of business strategy and technical execution. I've helped 100+ clients stay on top of their KPIs, shipped AI products that handle thousands of customer interactions every week, and today I'm the real-time nerve center of a financial services sales floor. I translate complexity into clarity.
Most analysts are strong on one side of the job: either they can build the technical infrastructure or they can communicate insights to stakeholders. I've spent the last several years getting serious about both, cleaning and validating data that feeds machine learning models, building BI layers that serve dozens of clients at once, engineering AI agents that handle thousands of customer interactions, and running real-time operations analytics for a large-scale service organization. Each project pushed me deeper into how data actually flows through a business and where the real decisions get made.
My edge is that I get both sides. I can write the SQL, build the data model, set up the pipeline. I can also sit with a non-technical stakeholder, understand the actual decision they are trying to make, and deliver something they can act on. Most people are strong on one side. I work in the middle, and that is where the high-value analytics work lives.
A few technologies I work with regularly:
I am not a dashboard-only analyst. I work the full width of the stack, from pulling messy data out of source systems to shipping the AI product that acts on it. Step through the five stages below to see how each one connects to the next.
01 / Ingest
Every reporting problem I have inherited started upstream: data sitting in a vendor export, a CRM nobody fully trusts, a spreadsheet three people edit at the same time. I build the pipelines that pull all of it into one place on a schedule, so the number on the dashboard Monday means the same thing it meant Friday.
I care about the unglamorous parts here, things like idempotent loads, sane naming, and schema changes that do not silently break a report six layers downstream.
02 / Model
Raw data is not useful to anyone until it has a shape. I design the models and definitions that turn scattered tables into a single source of truth: consistent grain, consistent naming, one definition of revenue instead of five.
That discipline is what let me move a full reporting platform to a new stack without anyone downstream noticing, other than that it got faster.
03 / Validate
A dashboard that is wrong once loses the room for good. Before anything ships, I test it: reconciling it against a source of truth, checking for the edge cases that quietly break averages, and building alerting so a broken pipeline gets caught before a stakeholder ever sees a wrong number.
This is the least visible part of the job and the part that protects everything else.
04 / Visualize
A dashboard is not the deliverable. The decision it enables is. I design views around the question a stakeholder is actually trying to answer instead of the tables that happen to exist, so people open it and act instead of squinting at it.
That is the difference between a dashboard that gets used once and one that becomes part of how a team runs.
05 / Activate
The best insight is one that does not need a human in the loop every time. I build the layer past the dashboard: agents and automations that read the data and take the next step themselves, whether that is answering a customer or flagging a staffing gap before it happens.
This is where analytics stops describing the business and starts running part of it.
Four habits show up in almost everything I ship, regardless of which tool is involved.
Before I open a single table, I want to know what someone is actually going to do with the answer. Monitoring a live operational metric only matters if it changes what happens next, so I build with the decision in mind first and the query second. That habit is what lets me dynamically adjust routing, staffing, and priorities in real time and keep a floor hitting its targets as conditions shift throughout the day.
I would rather spend an extra day getting a data model right than spend the next year explaining why two reports disagree. I design workforce forecasting models from historical patterns to project volume and staffing needs, and I build reporting layers meant to survive contact with reality: new data sources, edge cases, and the inevitable request for one more cut of the same numbers.
If I am doing something the same way twice, it becomes a script the third time. I engineered an AI chatbot on the OpenAI Assistants API that lifted customer satisfaction by 30%, and deployed a RAG-based agent handling 1,000+ weekly interactions that cut live-support workload by 20%. The goal is always the same: free people up for the decisions a model cannot make yet.
None of this matters if the person on the other end cannot use it. I publish recurring performance reporting that keeps leadership ahead of emerging issues, turn process inefficiencies into measurable capacity-planning improvements, and design dashboards simple enough that someone can open one without being walked through it first. Good analytics work ends in an action, not an attachment.
Swipe to explore
An interactive Tableau dashboard covering sales performance across KPIs, trends, and segment comparisons, built for fast, self-serve exploration.
Analyzed LinkedIn survey data in Power BI to uncover key trends, enhancing decision-making through interactive visualizations and data-driven storytelling.
Measured how TSP impacted WeGo bus schedule stability (Feb–May 2025). Cleaned and modeled ~1.2M rows in pandas and built Tableau dashboards for city transit planners.
Built end-to-end financial models for cash flow, investments, and loan amortization, with interactive dashboards and projections for retirement and business forecasting.
Sales Dashboard built in Tableau, interactive and embedded directly below.
I'm open to data analyst roles and freelance analytics work. Whether you have a question or just want to say hi, my inbox is always open.
Say hello