About
I've spent 12 years building products inside healthcare, from global consumer apps at…
Activity
2K followers
-
Alex Kramer posted thisThere's a pattern I keep seeing at practices that adopted an AI scribe early: the scribe becomes the finish line, when it was only ever supposed to be the on-ramp. Heidi (or Nabla, or Suki, etc.) listens to the visit and produces a clean transcript. Genuinely useful. Clinicians love it. And then the workflow looks like this: transcript, copy into Claude or ChatGPT to draft a care plan, paste into a Google Doc, export to PDF, send to the patient portal. Four manual steps, every visit, for a workflow that should take zero. The scribe didn't close the loop. It just moved the bottleneck one step downstream, and made it more visible, because now the copy-paste is the only manual step left in an otherwise automated visit. The practices that adopted a scribe first are, ironically, the ones most ready to fix this. They've already proven the team is comfortable with AI in the clinical workflow. The scribe was the trust-building step, not the destination. I built that next step directly: a care plan generator that fires the moment the clinical note locks and reads straight from it, no copy-paste, no separate tool. It lands in draft mode first, so the provider reviews it before anything reaches the patient, then one click sends it to the portal. Same information, same trust boundary, one less place for anything to get lost in translation. If your AI scribe made visits faster but your post-visit workflow didn't change, you didn't get the automation. You got a better dictation tool.
-
Alex Kramer posted thisBeen heads-down building the last few months - the good kind of busy, the kind where quiet on LinkedIn is a sign of progress. I'm back online and ready to bring some of it out from behind the scenes, starting with one of my favorite lessons from that stretch. I shipped a clinical note scribe feature, watched clinicians actually use it, and then watched trust erode fast: the notes it generated were often wrong, and fixing them meant real hands-on rework for the clinical team, visit after visit. Version one ran on a premise that felt right: more structure, more rigor. Parse the whole visit (history, assessment, plan, vitals) into clean structured fields. It looked great in a demo. But every mismatch between what the AI produced and what the platform expected meant an engineering fix, not a quick edit, so wrong notes stayed wrong for days at a time. That sent me back to the workflows themselves, and the answer that came back was better than what I started with: only a few things actually need hard structure and an audit trail: medications, labs/imaging/orders, referrals, problem list. Everything else, the clinical team wanted as narrative, notes that read like a clinician wrote them. The fix wasn't "structure more carefully." It was "structure only what actually needs it." So the rebuild started there instead of in the AI logic: an admin panel where fields and templates are configurable by the team itself, each one paired with a plain-language description of what to parse for and examples of what good output looks like, plus a way to flag intake questions that should just prefill straight from the form, no AI interpretation needed. The lesson I'm carrying forward: more structured data isn't automatically better clinical product. The right amount of structure is set by what needs proving, not by what's technically possible to parse. Getting that right made the product simpler, and the team happier with it. Next week: what changed once the review flow caught up to that same idea.
-
Alex Kramer posted thisFive months ago a client came to me with a problem I've seen before: strong brand, specialized clinical model, and a platform that couldn't keep up. Fragmented tools, performance issues, providers jumping between three systems to complete a single visit, critical workflows going offline. Sound familiar? They knew something had to change, and had big aspirations to deliver on their brand promise. They didn't know what it would cost to fix it. Here's where we landed: a fully custom provider and patient experience, built on Canvas Medical, delivered in two months(!) after a focused discovery sprint to make sure we were building the right thing. I built the core of it myself using the Claude Code studio on top of the Canvas API. The most fun was that this build was truly end-to-end: requirements gathering, systems and feature design, operational workflow design, feature development, testing, training, go-live. No engineering team. They're saving hundreds of thousands in platform costs and now have a product that actually fits how they practice. And they won't need to hire engineering for quite some time; I can stay on fractional to maintain and extend it as they grow. This engagement couldn't have existed a year ago. Not because Canvas didn't exist. Because a fractional PM who can take something from discovery to deployed — largely solo in two months — wasn't really possible before the current generation of AI tooling changed what one person can do. The founders who are winning right now aren't necessarily the ones with the biggest engineering teams. They're the ones making the right foundational decisions and finding people who can execute across the whole stack.
-
Alex Kramer shared thisShout out to Bunkerhill Health and Out-Of-Pocket for hosting last night’s rooftop health tech hang! Always great to trade notes with other folks on what we’re seeing in the space, how our product design decisions are upended by AI, and also where to draw the line on clinical decision-making. Our trivia team also did okay!
-
Alex Kramer posted thisMost healthcare product thinking runs through two lenses: what the patient experiences, and what the provider experiences. Roadmaps and design reviews optimize for one or the other. A third lens gets skipped almost every time: whoever has to operate and maintain the thing after it ships. Call it clinical ops, call it the admin: the person who has to update a rule or fix a workflow six months from now without an engineer in the room. That's always mattered, but it's urgent now. Claude Code makes it possible to build custom interfaces, configs, and business logic almost as fast as you can describe them. More of that logic is quietly moving from vendor config screens into code only an engineer (..or Claude) can read. Patient and provider lenses tell you if a feature should exist. The ops lens tells you if it survives contact with reality. Whether someone without a codebase can see what the logic does, change it, or turn it off. Skip that, and you've shipped a liability with a UI. The fix: treat the third lens as a design requirement, not an afterthought. Build the admin surface alongside the feature, not after someone gets stuck. Who owns that on your team right now?
-
Alex Kramer shared thisProud to be part of this one! Great piece from Nick Kavassalis and the team at Origin on what's possible when a lean team builds instead of buys. My part of it: Origin's clinicians had outcomes data: engagement stats, progress tracking, severity thresholds. It existed but wasn't accessible during patient visits. The "buy" option runs $250-400 per provider per year, which adds up fast across their 110+ providers. I built it instead: a Chrome extension on top of their EMR, done in six weeks for a few hundred dollars in Claude tokens. This is part of a bigger shift in what a product advisor can deliver now — less roadmap decks, more shipped product. Great to see it in action with this team.Alex Kramer shared this"Technology has been democratized. A small, passionate team can roll something out in weeks, not months. For the first time, we can incorporate the slight improvements and feature requests that came up time and time again. It's enabled us to build at a speed and cost that simply wasn't possible before." Learn about what our team is building in Nick Kavassalis's latest article. Special thanks to Alex Kramer and Angela Dobinsky for their input and contributions!How We're Building Our Way to Better Clinical CareHow We're Building Our Way to Better Clinical Care
-
Alex Kramer posted thisA few years ago, I led an evaluation of 14 EHRs before we chose one. Here's what I wish I'd known going in. The number that matters most isn't the licensing fee. It's the engineering tax; the hours per sprint your team will spend maintaining the integration instead of building product. We built a scorecard. Clinical workflow fit, API flexibility, billing compatibility, support responsiveness, and cost, but also: how much custom front-end work will this require? What's the migration risk if we outgrow it? We selected Healthie. We ran the full cutover across 12 clinics and nationwide telehealth in a single weekend: zero business downtime, ahead of our own best-case timeline. The old platform's licensing fee was never the real cost. The maintenance tax was — the engineering hours it consumed every sprint. Healthie eliminated both: a few hundred thousand dollars a year in direct platform costs, and a team that could finally get back to shipping product instead of babysitting an integration. With AI tooling now, the speed delta between a good and bad architecture decision is larger than it's ever been. The evaluation methodology is the part most teams skip because they don't have a template for it. Full scorecard at https://lnkd.in/ggPeBYBK
-
Alex Kramer posted thisI built a clinical-facing product in weeks. Patient engagement tracking, outcomes measurement, program building. Solo, with Claude Code. Technically ready to ship. Then I was asked to hold. Not because of a bug. Not because of scope. Because deploying it would require serious clinician training, and rushing that would kill adoption before it started. A few more weeks of prep would make the difference between a tool that actually changes how the team practices and one that sits unused in a tab. So the product sat, ready, while we prepared the humans. I used the time well, testing, beta users, feedback cycles. But it clarified something: AI has widened the gap between "technically ready" and "organizationally ready." The build side is faster. The human side isn't. Clinicians adopting a new workflow requiring behavior to actually change, build new habits, trust a new tool - moves at roughly the same pace it always has. That gap is new, and most early-stage teams don't have a name for it yet. The instinct when you can build fast is to ship fast. But at a clinical practice, you're not just releasing software. You're asking trained professionals to change how they work. That has its own timeline, and it doesn't compress. Ship-ready and adoption-ready are two different milestones. The second one is the one that matters.
-
Alex Kramer posted thisThe most diagnostic question I ask in any discovery conversation: "What do you copy and paste?" Every instance is a missing integration or a tool that doesn't do what the team needs. Staff know exactly where the friction is. Leadership often doesn't. What I find at almost every practice I work with: → Intake forms → EHR (field by field, manually) → AI scribe output → EHR chart note (copy-paste, every visit) → EHR data → care plan document (manually assembled) → Care plan → patient portal (emailed as a PDF) Ambient scribes are collapsing the documentation step — but that second arrow, scribe output → EHR, is still copy-paste at most practices because the integration isn't there. AI surfaced the bottleneck by removing the step before it. And that scribe isn't free. Standalone tools run at least ~$70/provider/month. At 10 providers that's $700/month. At 50 it's $3,500. It scales directly with growth, which means the architecture question gets more expensive the longer you wait to ask it. The options aren't obvious: your EHR may already include a native scribe (Healthie Scribe, athenaAmbient, Epic's AI scribe). Clinician satisfaction with the current tool might make switching painful enough to keep it. Or the cost of replacing it may be lower than you think once you run the math. But you have to run the math — most practices haven't. Every instance of copy-paste points at a decision that was never made. What did AI expose when it took away the step before the copy-paste?
Alex Kramer commented on a post
1w
Love to see this! And being a “human integration point” is a great place to start building a product.
Definitely interested in how you may harmonize the complete picture, especially when wearing an Oura ring isn’t practical (ie. I don’t wear mine when I’m lifting, only during cardio workouts).
-
Alex Kramer reacted on thisAlex Kramer reacted on this🚀 We’re growing at Mavida Health! As we continue expanding our mission to provide specialized, accessible mental healthcare for women throughout every stage of life, we’re excited to grow our team across several markets. We’re currently hiring: ✨ Business Development Associate – New Jersey ✨ Business Development – Southern California ✨ Business Development – Northern California ✨ Psychiatric Mental Health Nurse Practitioners (PMHNPs) – NY, NJ & CA ✨ Licensed Mental Health Therapists – NY, NJ & CA At Mavida Health, we’re building a team dedicated to transforming women’s mental healthcare through compassionate, specialized care across life’s many transitions. Interested in joining us? Explore our current openings and apply here: https://lnkd.in/gWdPE-We Know someone who would be a great fit? Share this post with your network! 💜 #MavidaHealth #NowHiring #WomensMentalHealth #MentalHealthCareers #HealthcareJobs #PMHNP #TherapistJobs #BusinessDevelopmentMavida Health | Online women's mental health therapy and prescribingMavida Health | Online women's mental health therapy and prescribing
-
Alex Kramer reacted on thisAlex Kramer reacted on thisMidi Health is taking women’s health to the next level! Midi Health is officially making its national retail debut at Ulta Beauty, bringing its clinician-formulated, science-backed supplement line to stores nationwide and online! For too long, women navigating perimenopause, menopause, and midlife health shifts have been left to decode crowded, confusing supplement aisles on their own. Midi is changing that trajectory by meeting millions of women right where they shop. From targeted cortisol support to brain clarity formulas, Midi's line is built on clinical insights from hundreds of thousands of patient visits delivering real efficacy, safety, and care tailored to women’s midlife biology. What makes this milestone especially meaningful is the opportunity to bring women’s health into everyday spaces connecting women with products designed around the biological changes that can occur during perimenopause and menopause, while complementing individualized clinical care. Congratulations to Joanna Strober and the entire Midi Health team on this exciting next chapter!
-
Alex Kramer reacted on thisAlex Kramer reacted on thisA personal update: after nearly eight years, I've decided to close this chapter at Willow, and today is my last day as CEO. When I joined the company in 2019, we had transformative wearable pump technology and a big ambition: to reinvent the breast pump category and make life better for new moms. Today, the wearable category we pioneered represents the majority of pumps sold, and millions of women are now pumping with more freedom, confidence and control. In the last few years, we've expanded our products to support moms across breastfeeding and postpartum needs, and with our acquisition of Elvie last year, we've grown into a global platform for maternal wellbeing with an exciting roadmap ahead. I'm proud of those milestones, but what I'll carry with me most is the impact we've had on the women we serve and the culture of this team, who pour their hearts into a mission that matters, united by a fierce belief that moms are worth fighting for. After many months of succession planning with our board, I'm thrilled to hand the reins to Carolin Archibald, Willow’s next CEO. Carolin knows this category deeply, cares about it personally, and leads generously. Willow could not be in better hands. Thank you to all of Willow’s investors, partners and champions, and to every mom who has trusted us with her feeding journey. Leading Willow has been the privilege of my career. I’ll stay connected as an advisor and remain a passionate advocate for women’s health. And I will forever be cheering this team on. ❤️
-
Alex Kramer reacted on thisAlex Kramer reacted on thisVery excited to announce some big news! Elmhouse Learning, an Outsmart subsidiary, has acquired Elmbridge University. I’m thrilled to now serve as Elmbridge University’s Executive Chairman. It’s a pivotal time to innovate in higher ed. The need and the opportunity are tremendous. More in the comments.
-
Alex Kramer reacted on thisAI Charting within Healthie has gotten a whole lot more powerful. We see millions of minutes flow through Scribe each week, and this is now supplanted by a broader suite of solutions that follow a session from beginning to end. Specifically: ✅ Scribe: Chart Notes drafted within your specific template, per Appointment Type ✅ Encounter Summaries: Automatically generated a concise clinical summary upon the note being locked ✅ Scribe Tasks: Follow-up tasks drafted from session transcript, commonly used for referrals, lab orders, care plan assignments ✅ Chart Summaries: Key insights and Patient Synopsis generated pre-session within Patient Profile; updates in real-time based on new information presented in Chart Note [in Beta, reach out to sign up] Link to learn more in comments!
-
Alex Kramer liked thisAlex Kramer liked thisSchool was closed today for a professional development day, but work… was still work. 🥴 (truth be told… I forgot to request it off!) My son did art beside me before a meeting and proudly showed it to my coworkers. When I started to apologize, the response was, “No need to apologize.” Later, we took a clarity break and went for a walk. We looked for grasshoppers, stopped at a patch of wildflowers, and noticed the light coming through the trees. This is what workplace flexibility looks like to me. It’s the room to care for your child when school and work schedules don’t align, and coworkers who make that room feel normal. That kind of support matters to parents and families. Today, it meant I could be present for my work and for my son 🫶 Let’s make this the new normal!!
-
Alex Kramer liked thisAlex Kramer liked thisToday EliseAI is announcing a $350 million funding round at a $4 billion valuation, led by Andreessen Horowitz (a16z) and Bessemer Venture Partners, with participation from Ontario Teachers' Pension Plan, Sapphire Ventures, and Navitas Capital. Over the past few years, Elise showed that AI can change the economics of a foundational industry. 1 in 5 multifamily apartments in the US run on our products. Those units see 2pp higher occupancy. And their operators see up to 20% higher NOI. It points to a larger truth about the intelligence age: this extraordinary technology can raise the standard of ordinary life for everyone. But for it to do this, someone has to do the work of building AI for the hardest, most important industries in the world. So that, as intelligence gets cheaper, life does too. And the benefits of AI reach everyone. This round gives us the resources to go deeper and work faster. I wrote about everything we’ll accomplish in housing and why we’re bringing AI to essential services industries. Thank you to our investors, our customers, and our team. There’s so much to do, and I’m excited to keep building together.
-
Alex Kramer liked thisAlex Kramer liked thisBHT really does feel like home! 💜 A few highlights from the past week: - Got to catch up with some of the people I’ve been lucky enough to work with: Emma Sugerman, Dawn Androphy, Connor Gordon, Amy Brandli, Alexander Mueller, Aaron Reichlin, MD, Brian Yam - Spent two weeks traveling with my amazing cousin (and coworker!) Paula Pereira and got to join Macky Rocabado Gay on her first Light-itt-it trip. Both absolutely killing it! - Had a karaoke moment with Erez Lirov 🎤 - Got to see some Nashville bands that absolutely blew me away! 🤠 Reconnected with so many amazing people: Micah Friedland, Tom Kim, Dr. Ashish Mandavia, MD, Carmine Di Maro, Nick Neral, Jessica Holton, Erica Jain, Caseley B Pinto, Greg Milnarik, Noah Goldfarb, Julia Morgan Knight, Derek Strauss, Edmond Coku, Ignacio (Iggy) Handal, Damayanti Dipayana, Jaime Calderon Goff, and met so many new people doing exciting work who I’m hoping to add to my “friends of BHT” list :) Mike Lifshotz, Rich Sneider, Douglas R. Polster, Ph.D., Alexander Alvarado, Psy.D., AK Azaiez, Danny Gladden, LCSW, MBA, Guedis Cardenasl Humberto Buniotto, and so many more!! Already looking forward to the next one! #BHT2026 Behavioral Health Tech
Experience & Education
Licenses & Certifications
Honors & Awards
-
Cum Laude
Cornell University, Information Science Department
Awarded Cum Laude for research in Information Science.
Languages
-
English
Native or bilingual proficiency
-
Spanish
Professional working proficiency
Other similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content