Sid Bhatt

Today we officially kick off our $2M pre seed round for Salubrum (Backed by Forum Ventures), supported by Forum Ventures, Coeus Collective Ventures, and Draper House Americas. Healthcare companies spend over $30B a year trying to acquire patients. The problem is not media. The problem is timing. Every time a patient approaches a healthcare decision a moment opens. Which doctor to see. Which procedure to pursue. Which treatment to start. For the companies involved in that decision, knowing that moment exists is incredibly valuable. Salubrum (Backed by Forum Ventures) is building the intelligence layer that maps those decision moments across the U.S. patient population. Our system aggregates large healthcare datasets covering 300M+ patient identities and uses them to predict when individuals are approaching important healthcare decisions. That signal can then be used by providers, pharma companies, payers, and other healthcare organizations trying to reach patients at the right time. Our first wedge is patient acquisition for healthcare providers and brands, but the underlying intelligence layer has much broader applications wherever patient behavior influences economic decisions. Over the past months we have quietly built the foundation. • Early pilots validating our patient intent scoring system • Distribution partnerships forming across healthcare platforms • Backed by Forum Ventures, Coeus Collective Ventures, Draper House Americas, NEXT Canada and Metta World Peace Starting tomorrow we will be meeting investors as part of the Forum Ventures showcase. If you are an investor interested in building with us get in touch. Calendly: https://lnkd.in/e5j-VvGC Investor Memo: https://lnkd.in/e5HyemAp

95

6 Comments

With RL techniques like GRPO, it’s become easy to fine-tune small models to outperform foundation models on vertical tasks. We’re open sourcing Osmosis-Apply-1.7B: a small model that merges code (similar to Cursor’s instant apply) better than foundation models. Links in comments! Using foundation models for high specificity, low complexity tasks like code merging can be overkill - instead, it’s better to use a specialized model instead for higher performance and latency. We fine-tuned the model on 100K real-world code commits where the model needs to merge a code edit snippet into an original code snippet. On the test dataset (subset of CommitPackFT), foundation model performance ranged between 0.77-0.93 reward score. In comparison, Osmosis-Apply-1.7B achieved a 0.98 reward score - while also being 3X-5X more affordable than the next cheaper model AND ~10X faster! Why this matters: the best AI agents use multiple models to ensure task specialization and reliability at scale - on top of cost and latency benefits as well. Links to the model and full write-up (with reward function, hyperparameters, etc.) are in the comments - reach out if you’re interested in using reinforcement learning!

240

18 Comments

Big day! Nitra has raised $187 million in combined financing as we build the AI-native operating system for healthcare practices. We're also announcing that Dr. Richard Park, founder of CityMD and healthcare legend, will be joining Nitra’s Board. But the real story of Nitra starts with a question. Why does the largest industry in America (to the tune of $5.9 trillion a year) still run on fax machines, spreadsheets, and 1-800 sales reps? The doctor who just operated on you is probably spending hours every day on phone holds, paper invoices, and reconciling accounts across six disconnected systems. Nobody built the operating system for the business of healthcare. So we did. In the past year alone, Nitra has grown quickly: • $1B+ annualized processing volume across 700+ clinics and thousands of physicians • 740%+ revenue growth in 2025 to $33M+ in annualized revenue • $9M of biopharma and surgical supplies processed in a single day That’s what it looks like when software becomes part of the infrastructure of healthcare. Doctors are not just service providers. They are the backbone of every community, yet many are forced to act as part-time CFOs, insurance specialists, and procurement officers just to run their practices. They deserve software that works as hard as they do. Huge thanks to our team, our investors, and the physicians building with us every day. And to Jonathan Chen, who’s been building this with me since day one. Appreciate Fortune and Catherina Gioino for the story: https://lnkd.in/gtHyn2Wg To read more about the Nitra story check out our blog here: https://lnkd.in/giHwVry2 To apply for open positions: https://lnkd.in/gePZdYRt If you’re a physician or operator who wants to modernize your practice, or if you want to help us build the future of healthcare, we’d love to hear from you.

1,134

198 Comments

I’m so proud that Hyphen Capital has been a backer of Tim Hwang and Jonathan Chen since the beginning of Nitra. They had big ambitions to modernize the archaic back office of healthcare practices from fax machines to AI. It’s been amazing to see their growth from processing zero transactions to over a billion dollars and they’re just getting started.

40

3 Comments

TheFounderVC has a truly compelling team - entrepreneurs and operators that have scaled from 0-1 and to exits across a variety of markets - some of whom are fellow classmates. They have built strong founder and VC networks over the last 25 years in tech and are now leveraging that access and experience on behalf of their portfolio. It clearly resonates with founders, as their early portfolio is truly compelling.

19

AI and Health Tech in Focus   AI Tool Used to Help Predict Postpartum Depression Researchers at Harvard University’s Massachusetts General Hospital have developed an AI tool that helps predict the risk of postpartum depression (PPD) before discharge from the hospital. Approximately 15% of women suffer from PPD which is responsible for suicide and self-harm and about 10% of all pregnancy related deaths. The researchers trained the algorithm based on about 29,000 mothers using electronic health records data including medical history, medication data, age, race, ethnicity, education, marital status, insurance type, number of births, mode of delivery, number of prenatal visits, length of delivery stay as well as an existing but more modest predictability score for PPD called EPDS. Incorporating this data resulted in an ability to predict more than three times the baseline risk of PPD. Using this model can help care teams identify those at risk for PPD and provide resources and support services such as referral to social workers, psychotherapy, psychiatry or medical therapy before they have an impact on the mother, child and their families. VC Investors are always adjusting their "Play Book" with a variety of options for investing. Mikal Ventures is one of those alternatives, offering curated solutions to real world problems. Reach out if you have questions on how we can be of assistance and bookmark https://mikalventures.com/ for additional insights and information.

1

Health tech startups are scrambling to adapt as labor shortages and pricing pressures loom large. With EHR giants like Epic calling the shots, how are founders supposed to keep their heads above water while navigating this chaotic landscape? #HealthInnovation #DigitalHealth https://lnkd.in/gYbCjspE

1

FANTASTIC news at Ubiquity Ventures-backed Ketryx! Co-founders Erez Kaminski and Jan Pöschko have just closed a new $39 million Series B financing round. Ketryx offers an AI-powered compliance platform that lets people all over the world design, develop, test and deploy highly-regulated products at a fraction of today's time and cost. It's an AI-native platform that lets you interconnect different IT systems with AI agents to accelerate the rate of safety-critical product development. Three of the world's top five MedTech companies already trust Ketryx to build safer software faster—and we're just getting started. Glad to have partnered with Ketryx from the very beginning on behalf of Ubiquity Ventures alongside E14 Fund (Calvin Chin & Habib Haddad) and bring on Lightspeed (Guru Chahal) and 53 Stations (Kevin King) at our Series A 18 months ago. Watch more below!

27

4 Comments