Started in operating rooms, not conference rooms. Cardiac devices at St. Jude, wearables at Under Armour and Owlet, industrial IoT at Ribbiot — fifteen years shipping systems that have to be right the first time.
I build products that connect people, data, and devices. Fifteen years in — starting as a clinical specialist in operating rooms at St. Jude Medical — I have shipped connected hardware and cross-platform software, applied AI to real user problems, and led squads through ambiguous 0→1 work and iterative scale-ups.
Previously at Owlet Baby Care and Under Armour, and currently at Ribbiot, I partner with firmware, mobile, and data science teams to turn complex signals into clear, trustworthy experiences. My approach mixes user insight with analytics and modern tooling to deliver measurable outcomes.
Austin, TX
15+ years
Medical devices, connected products, AI-augmented SaaS
Healthcare, fitness, industrial operations
A walk through three products I led — from FDA-cleared wearables to connected footwear and an IoT platform built from zero. Measured by outcomes, not ornament.
Ribbiot Operations Platform. 0→1 IoT + SaaS platform for crane and rigging operations. Connected telematics, sensors, and mobile workflows; shipped the first integrated product in three months and drove 3× DAU growth in a tech-averse industry.
Owlet Dream Sock. FDA-cleared wearable (SaMD) translating physiological signals into parent-facing insights. Improved NPS from 10 → 60 in one year; led cross-platform delivery across firmware, iOS, and Android.
MapMyRun & Connected Footwear. Sensor-driven coaching across iOS, watchOS, and embedded footwear. Helped runners go 7% farther and 3% faster; scaled to 1.25M+ device activations and 5M+ MAU.
Remote
Remote
Austin, Texas
Austin, Texas
Austin, Texas
Austin, Texas
Austin, Texas
Austin, Texas
Five years in labs and hospitals before the first PM role.
BS + MS, Biomedical Engineering
Cal Poly, San Luis Obispo — 2011
Thesis: Ischemia Impairs Vasodilation in Skeletal Muscle Resistance Artery
Was headed for the MCAT until St. Jude called. The thesis worked for both paths — what actually stuck was the method, and knowing that 95% of real research never gets written down.
The daily kit — what I reach for to discover, build, and ship.
Four principles I keep coming back to — earned in the OR, sharpened in product rooms.
Make it visible.
I started my career guiding cardiac surgeons in the OR. I learned young that the room only changes when the room sees it. Every PM meeting since has been the same dynamic with lower stakes.
— 5 years in ORs and physician officesSell the feeling.
When Owlet lost its FDA clearance, we couldn't ship the data anymore. So we rebuilt peace of mind through different channels — new alerts, new defaults, reset expectations. Same feeling, different mechanism.
— NPS 10 → 60 at OwletPrototype before you PRD.
Learning beats documenting. The PRD is better when the open questions are already answered. The prototype is the spec; the doc is the cleanup.
— ~80% setup time reduction on AI formsGive the problem, not the ticket.
I give engineering the problem and the context; they own how it gets built. The team ships faster, the solutions are better, and I'm solving the next problem instead of writing the current ticket.
— AI-augmented context → engineering-owned deliveryI used to describe the thing; now I build the thing. This is some of what that's looked like.
Every Gong call, run back through a PM-shaped summarizer, stacked with a reporting layer, a chatbot over the data, and an MCP server so every other tool in my stack can ask questions of it too.
Sales notes are written by sales, for sales. The product signal was in the recordings; it just needed someone who cared about a different question to go get it.
Feed it a PDF, it generates a form data model. One of the first real AI features I shipped at Ribbiot, with the team.
Twelve months ago this saved our one-person CS team real hours. Today I can out-prompt it manually in a chat window. Leaving it in here because watching your own work get lapped by the frontier is part of the job now.
Cursor as command center, MCPs as the arms and legs. Reaches into Linear, Notion, GitHub, Gong, Zendesk, and TestRail, with personas and competitive research on standby. Drafts specs, PRFAQs, help docs, test cases, marketing copy — whatever the moment calls for.
PM work is ninety percent moving the same idea between ten tools that don't talk. I got tired of being the integration layer, so I built one.
A cooking app that starts from your fridge, not a search bar. Built solo in Cursor and Claude on Expo, live on iOS with Android right behind it. Tell it what you have; it gives you two or three things to cook tonight.
Five years ago a PM could not ship a native mobile app to two stores alone. I did. That's most of why this entry is here — Cookfor is also the product I most want to exist, but the fact that one person with the right tools can put a real app in front of real users is the whole point of this page.
Also, because I could: a birthday site for my 5-year-old and a half-marathon training tracker. Neither is important. Both exist because the cost of building dropped to near zero, which is sort of the whole point.
The fastest way to reach me is email. LinkedIn works too.
kstruthe @ gmail.com