shawnnext.ai

Shawn M. Brown

Exploring AI — building tools, experiments, and ideas at the edge of what's next.

Projects

SherlockEQ macOS doesn't ship a per-ear parametric EQ. SherlockEQ is one. It taps system audio, splits the chain by ear, and gives you the controls to investigate a tinnitus tone, shape an audiogram into compensation, and refine sound to taste. No virtual driver. No kernel extension. No paywall.
Cullapse Cullapse is an inbox triage tool — not an email client. It connects to your existing accounts, fetches only headers, and never touches a server we own. No account creation. No subscriptions. No ads. No tracking. No investors — which means no one is pressuring us to monetize your attention, harvest your data, or grow at your expense. We answer to you, and only you. Processing is local. Privacy is the whole point.
Soon
BeatCheck Web A bare-bones, self-hosted RSS reader with a Claude-powered AI pipeline that scores, clusters, and curates news for a single user. Built with FastAPI + HTMX + SQLite. Designed to become an agentic tool for turning news into newsletters, websites, podcasts, or other outputs over time. Inspired by and partially derived from Leo Laporte's BeatCheck.
Shopper Intervention ML System An end-to-end machine learning system that predicts which e-commerce sessions are unlikely to convert, enabling timely promotional interventions to recover potential lost sales. Features a FastAPI backend trained on 12,000+ sessions with champion/challenger model selection, a Streamlit dashboard for interactive scoring and analysis, and automated retraining with MLflow experiment tracking.
Tylenol Dosing Calculator This calculator provides general dosing guidelines based on standard recommendations. Always consult with a healthcare provider before administering medication.
Hospital Census & Staffing Dashboard The dataset is an Excel workbook with one month of census and nurse staffing data for four hospital units (MedSurg, ICU, NICU, and ED Observation). It includes a daily unit-level summary tab, an hourly census tab, and a “Mitigation Playbook” tab that defines the Red/Yellow/Green thresholds, mitigation actions, and planned staffing by day of week. Using these inputs, the requested output is a 5-day forecast tool that predicts each unit’s average daily census and related nurse staffing needs, then assigns census, staffing, and overall (worst-case) color status per unit—and reports the predicted 5-day census and color coding for each unit.
DIGit An AI-powered document analysis tool that uses Claude to break down content into Data, Information, and Guidance using the structured DIG framework.
Next Project Brief description of what this project does
Soon