Projects & work samples
Things I've built
Working software, live on the internet — not slideware. Every project below runs as a live demo you can open right now, with the full source on GitHub. The write-ups tell you what each one does. How — and why it works — is a conversation.
★ Featured workForecastingAnalyticsPython · Streamlit
Pipeline Command Center: seeing the quarter before it happens
An executive sales analytics dashboard that turns raw CRM data into answers: how much revenue is actually going to land, where the pipeline is quietly thinning, and which loss reasons keep repeating. The forecast isn’t a hopeful line on a chart — it’s backtested against history and shown with honest confidence intervals.
The interesting part is what it flags that most dashboards miss. Open the demo and see if you can spot it.
Ask me how it’s built →
Machine learningExplainabilityPython · XGBoost
Lead Scoring Engine: which leads convert — and why
A machine-learning engine that scores every lead the moment it arrives — and, unlike most models, explains itself: for every score, it shows exactly which signals pushed that lead up or down. In the demo, move the sliders and watch a lead go from ice-cold to sales-ready in real time.
Here’s the hook: the signals that matter most are not the ones most teams chase. The demo shows you the what. The how is worth a call.
Ask me what the model found →
Original conceptOptimizationPython · ML
Superfan Revenue Engine: RevOps for an industry that’s never heard of it
Revenue-operations thinking applied where nobody applies it: independent music. The engine scores fans by expected value the way sales teams score leads — then an optimizer plans a tour that beats the obvious big-city routing on the same budget. In the demo, move the budget slider and watch the plan reshuffle itself.
Why it consistently outperforms the naive plan is the part I don’t write down. Ask me.
Ask me why it wins →
