BIOMEDICAL ENGINEER
PRESS ENTER TO START
Biomedical Engineer
Two years watching devices fail patients. Now I test them, validate them, and build tools so they don't.
Targeting validation, applications, R&D, and human factors roles in SoCal medtech.
biomedical engineer
01 / About
Every medical device has to prove it works and prove it's usable before it touches a patient. The testing. The validation. The clinical insight that connects engineering to real use. Whether I'm writing a test protocol, running a usability study, prototyping a sensor system, or training a team on a new tool, the through-line is the same: I understand what happens when the device meets the user, because I spent two years watching it happen.
| Parameter | Value |
|---|---|
| Clinical observation hours | 400+ |
| Device projects built and tested | 6 |
| Regulatory frameworks | IEC 62366-1, FDA 21 CFR 820, EU MDR |
| Degree | M.S. Biomedical Engineering |
| Tools built | UseTrace HFE, PhotoCull AI |
02 / Skills
Write IQ/OQ/PQ protocols and design risk-based test strategies (IEC 14971, FDA 21 CFR 820). Lead design verification, document deviations, and drive CAPA workflows. Build test traceability matrices tied to user needs and design inputs.
Own end-user training and support: live onboarding sessions, user guides, and troubleshooting docs. Translate regulatory requirements and engineering constraints into clear workflows. Collect field feedback and write feature requests for R&D.
Lead sensor system design end-to-end: requirements, schematic, firmware, validation. Signal processing pipelines (filtering, feature extraction, classification). Rapid iteration with MATLAB/Python simulation before hardware. Motion capture, EMG, acoustic, and pressure sensing.
Design and run formative usability studies. Classify observations under IEC 62366-1. Write HFE reports. 400+ hours of clinical observation watching real users interact with real devices.
03 / Story
400+ hours of clinical observation watching patients interact with medical devices. I saw the disconnect. Engineers designing in isolation, safety reports that missed real-world use patterns, and a review process that hadn't changed in decades.
So I went to Stevens for my M.S. in Biomedical Engineering. Design verification, usability evaluation, regulatory strategy, the whole device lifecycle. Got my Google UX Design Certificate because the gap between what engineers build and what patients actually experience is where most risk lives. Then I started building UseTrace HFE, an AI tool that classifies safety observations under IEC 62366-1 so engineers spend less time on paperwork and more time on the test plan.
Two years watching the problem. Then I went to grad school to fix it.
04 / Projects
Formative usability studies generate 200+ free-text observations per session. Classifying them under IEC 62366-1 use error categories takes 2-4 weeks with low inter-reviewer consistency. Missed classifications can hide safety signals in HFE reports.
Multi-label classification (severity, type, device subsystem) with confidence scores and audit trails. Engineers validate AI-generated drafts instead of coding raw text. Cuts classification time to 3-5 days and maps directly to HFE report deliverables.
| Specification | Performance |
|---|---|
| Classification time | 2-4 weeks manual → 3-5 days with UseTrace |
| Data capture | Structured observation capture during live sessions |
| Report output | IEC 62366-1 formatted reports in minutes, not weeks |
| Regulatory alignment | Maps observations to IEC 62366-1 use error categories |
| Adoption | 50+ engineers trained across 4 medtech companies |
UseTrace launched to 4 medical device companies. Within the first month, 50+ HFE engineers adopted it. When one team reported that observation categorization was too slow on older laptops, I profiled the React render pipeline, found unoptimized re-renders, and shipped a fix in v2.1 that improved speed 3.2x. That's the cycle I bring to every project: ship, listen, fix, ship again.
My HFE process follows IEC 62366-1 from use specification through summative evaluation. Every project starts with intended users and use environment, then builds task analysis, identifies use errors, and validates through formative testing with real participants. This isn't optional documentation. It catches what untested design misses.
See the full methodology and results in each case study below.
05 / More Work
Free, local-only photo quality scorer. Analyzes sharpness, faces, duplicates, and composition across 10 dimensions. Built it because I got tired of paying $10/month for Aftershoot.
M.S. capstone. A device concept that estimates air leak severity by comparing ventilator-delivered tidal volume with chest tube outflow. No existing commercial solution. Currently in design control phase.
Wearable electromyography concept that detects voluntary hand movement and suppresses Parkinsonian tremor via pneumatic actuation. Full Python simulation built from scratch.
Clinical research analyzing gait biomechanics to predict fall risk. Built a Python pipeline from scratch to clean Qualtrics data, compute reliability metrics, and run classification models.
Free progressive web app that uses pose estimation to coach exercise form in real time. Gamified PT exercises because I saw patients skip rehab when it got boring. Full circle from those clinics.
06 / Toolkit
Not a LinkedIn skills section. These are tools I've used to build things that work.
| Domain | Capabilities | Standards / Tools | Depth |
|---|---|---|---|
| Testing & Validation | V&V protocols, test methods, IQ/OQ/PQ, risk-based strategies | IEC 62366-1, FDA 21 CFR 820, EU MDR | Coursework + capstone application |
| Regulatory & Quality | Design controls, CAPA, computer software assurance | FDA 510(k), ISO 14971, 21 CFR 11 | Capstone + coursework, applying to real device workflows |
| Design & Human Factors | Formative usability studies, use error classification, observer analysis, HFE reports | Google UX Certificate, IEC 62366-1, ISO 9241 | 400+ hrs observation, UseTrace deployed to 4 companies |
| Software | Web apps, data pipelines, client-side ML, APIs | JavaScript, Python, TensorFlow.js, Firebase, Git | Multiple production apps (PhotoCull, PhysioRep, UseTrace) |
| Machine Learning | Classification, cross-validation, signal processing, statistical analysis | TensorFlow, scikit-learn, pandas, MATLAB | Coursework (BME 571) + PhotoCull ML pipeline |
07 / Journey
Kinesiology to biomedical engineering. Clinical observation to device design. Each step got me closer to the intersection of users, technology, and patient safety.
Watched devices fail patients. Decided to fix the problem from the engineering side.
Targeting roles where clinical insight and engineering overlap.
07.5 / Off the Clock
08 / Skills Map
Hover to explore which skills each project uses
| Skill | Pneumothorax | UseTrace | PhotoCull | EMG | Fall Risk | PhysioRep |
|---|---|---|---|---|---|---|
| Python | ||||||
| MATLAB | ||||||
| Signal Processing | ||||||
| Computer Vision | ||||||
| Machine Learning | ||||||
| FDA Design Controls | ||||||
| IEC 62366-1 | ||||||
| ISO 14971 | ||||||
| V&V Protocols | ||||||
| Usability Testing | ||||||
| React / JavaScript | ||||||
| Statistical Analysis |
08.5 / My Criteria
Most new grads position themselves as the ones being evaluated. That's half the story. Here's what I'm evaluating on my side, so we both save time.
If this list describes your company, my resume is one click away. If it describes half your company and you're working on the other half, I want to hear about that too.
09 / Contact
Device testing, usability studies, regulatory submissions, tools that help your team move faster. If your team builds devices that reach patients, that's where I want to be. SoCal preferred, remote works. Strong opinions about test protocols, decent taste in coffee.
(424) 421-9173