Projects

Served as end-to-end product manager for PredictionHealth Insight, a React-based analytics and compliance platform for clinical teams. Scaled to thousands of users across 200+ accounts. Owned the full PM lifecycle: customer discovery, user story writing, engineering requirements, sprint tracking, QA, and post-launch support. The platform withstood national-scale compliance audits, outperforming manual audit processes and helping clients avoid thousands of dollars in regulatory fines. Acted as the single point of contact between clinicians and the engineering team, ensuring feedback from the field continuously shaped the product roadmap.

Served as clinical subject matter expert on Sidekick, a first-of-its-kind ambient AI scribing tool purpose-built for physical and occupational therapy. Unlike dictation-based tools, Sidekick listens to natural clinical conversation and generates documentation automatically. Contributed product direction that shaped the ambient listening model, reducing clinical documentation time from hours to minutes. Scaled to thousands of users with high adoption rates. Labeled training data, validated model outputs against real-world clinical standards, and translated field feedback into prioritized feature requests that guided model development and roadmap priorities.

Built a consumer Android app from concept to production as a solo product owner — using Claude and Gemini as development partners while owning every product decision. Chore Quest solves a real gap in the market: most chore management apps are overcrowded with features that are too complex for young users and more than most families need. Designed a simplified, kid-friendly experience focused entirely on the features that matter — easy enough for children of any age to use independently, and flexible enough for parents to configure on the fly. Took the product through the full lifecycle: problem definition, requirements, design decisions, AI-assisted development with Gemini, Firebase real-time database integration, testing, and deployment. The app is now live with real users, generating organic interest from early adopters, and launching on Google Play Store this month.

Built a personal web application to solve a problem that most fitness apps consistently miss: the space between "too many features" and "not enough substance." Summit to Surf Fitness Tracker provides a clean, stable environment for planning workouts and tracking progress over time — without the noise that makes most fitness tools frustrating to use consistently. Built solo using Claude and Gemini as development partners, backed by a Firebase real-time database, and deployed as a web application. The project reflects the same product thinking applied professionally: identify a real gap, define the minimum feature set that solves the actual problem, and build something stable enough to use every day.

When Prediction Health was acquired by Prompt EMR, led the operational and data integration side of the transition. Documented and translated all internal processes, automation workflows, and data pipelines into the new organization's systems — ensuring continuity across customer success, support, sales, and operations. The priority was getting the data aligned: migrating, reconciling, and rebuilding reporting infrastructure across two organizations using BigQuery and Data Studio. Executing that foundation correctly is what delivered full operational visibility from day one.

Inherited an Asana environment that had grown too basic for the team's pace of work — flat task lists with no structure for tracking work through a lifecycle. Redesigned the entire workflow architecture around agile principles: built a tagging system, defined task lifecycle stages, and reorganized projects so work could be followed from intake through completion without falling through the cracks. Designed sprint dashboards used in weekly sprint meetings to surface blockers, progress, and priorities in real time. The result was a program management system that could grow with the team — turning a basic task list into a dynamic, queryable structure that gave leadership and individual contributors shared visibility at every stage of active work.

Led end-to-end ownership of an online scheduling platform rollout across 36 clinic locations — from initial vendor evaluation through post-launch support. Vetted vendor options, managed account setup and configuration, worked through platform bugs directly with the vendor team, and authored all supporting documentation for clinical and administrative staff. Coordinated across clinic directors, operations leadership, and front desk teams to align timelines and drive adoption at every location. Provided structured aftercare to resolve post-launch issues and ensure the platform stuck. The initiative resulted in a significant increase in patient volume and a measurable decrease in unfilled availability on therapist schedules — turning previously unbookable open slots into booked appointments.

Identified a gap in how new clinical staff were introduced to the internal applications they were expected to use from day one. Designed and built a structured learning management system with step-by-step walkthroughs for each core platform — not just general orientation content, but detailed, referenceable tutorials that walked users through real tasks in the actual tools. Staff could revisit the content as needed rather than relying on colleagues or recalling training from memory. The result was a measurable decrease in follow-up questions and a significant reduction in the frustration new hires experienced navigating unfamiliar applications without guidance.

The existing FOTO outcomes reporting process was outdated, difficult to navigate, and kept data out of reach for the clinical and operational leaders who needed it most. Redesigned the entire reporting pipeline using Power BI and Power Automate — pulling outcomes data from 36 clinic locations and transforming it into accessible, visual dashboards built around a system staff already knew. Instead of requesting reports or struggling with the source platform, performance data was surfaced automatically where leadership could act on it. The result was an improvement in survey and outcome collection rates across the network and a stronger, more reliable data foundation for drawing meaningful clinical and operational conclusions.

Designed and built a company-wide BI dashboard giving leadership real-time visibility into therapist performance metrics across all clinic locations. Beyond passive reporting, built a two-way communication layer directly into the dashboards: managers could trigger performance messages automatically when key metrics crossed defined thresholds, or send them ad-hoc based on what they were tracking in a given sprint. The automation handled delivery, timing, and logging — no additional operational headcount required. The result was more consistent, data-driven communication with therapists across the organization, giving managers a scalable way to stay connected to their teams without adding staff to support it.

Embedded with the outreach sales division to understand their core challenge: they were meeting with referral sources in the field without access to the data that would make those conversations count. Built a mobile app that put referral data, provider history, and pipeline status directly in reps' hands while they were on the road. Backed by a data pipeline connecting HubSpot and the clinical EMR, every rep could walk into any conversation with current, accurate information. The result was an increase in referrals and more productive discussions with referral sources — because the sales team now had everything they needed before walking in the door.

At Prediction Health, new client implementation, provisioning, and onboarding were tracked entirely in spreadsheets — manual, error-prone, and impossible to scale. Mapped the full process end to end, identified every step that could be systematized, and automated 90% of it using Google Forms, Google Sheets, Zapier, Asana, and Slack. Intake triggers kicked off provisioning workflows automatically, tasks were assigned and tracked in Asana without manual handoffs, and Slack notifications kept the right people informed at every stage without requiring someone to chase status. As the client base grew, connected the pipeline into BigQuery and Google Data Studio to add reporting and performance analysis on onboarding trends. The result was a scalable implementation system that absorbed growing client volume without additional operational headcount — and removed the human touchpoints where mistakes were most likely to happen.

Personally responsible for onboarding every new employee at IRG to the full suite of technical applications and hardware — covering pre-session prep, on-site walkthroughs, and follow-up troubleshooting. After running the process manually across multiple cohorts, identified that the majority of technical setup steps were repetitive and automatable. Rebuilt the onboarding workflow so that 80% of background processes — account provisioning, software installation, and system configuration — ran automatically before the employee arrived. In-person time shifted entirely to hands-on walkthroughs, live questions, and real troubleshooting. New hires spent less time watching progress bars and more time with someone who could actually help them get productive.

A marketing employee was spending 10+ hours every week manually reviewing batches of returned mail — reading each letter, identifying the bad address, and correcting the record in HubSpot or the clinical EMR. Built an image-to-text automation that ingested returned letters, extracted contact information, and cross-referenced it against the contact database to identify and flag outdated records automatically. When a batch came in, the system handled the identification work — eliminating the manual lookup and correction loop entirely. The result was 10+ hours returned to that employee each week to focus on higher-value marketing work.