CapitalG

Doctors are burned out. Patients don’t understand the advice they’re given. And the U.S. healthcare system is drowning each year in $1.5 trillion – with a “T” – in administrative costs. Abridge CEO Shiv (Shivdev) Rao thinks he can help. A cardiologist who still sees patients, he co-founded Abridge in Pittsburgh in 2018 to build the “AI operating system” for healthcare. Abridge's software automates the clerical work of a visit, from notes to billing, and increasingly assists with making clinical recommendations, too. "We think healthcare is the biggest opportunity for AI," Rao says. "We want to be deflationary, we want to save time, save money, and save lives." Valued at $5.3B, Abridge serves more than 300 health systems that reach 250M Americans, and recently partnered with Eli Lilly and Company and NVIDIA. But it faces plenty of potential competition, from giants like Microsoft, Epic, and Palantir Technologies, to VC-backed rival Ambience Healthcare, and maybe even the AI labs like Anthropic and OpenAI. On The Upstarts Podcast, Rao talks about how an early bet on AI got Abridge going, before ChatGPT; why reaching scale in healthcare requires years of “eating glass” and “kissing frogs;” and why he thinks Abridge can help doctors and nurses feel like “superheroes.” Plus, he shares his Upstart Moment: A “YOLO” decision to take on Microsoft head-to-head. This season of the show is presented by Rippling 🫡 https://lnkd.in/gMVFj7Ub

Pumped to share that CapitalG is backing K2 Space Corporation in their $500M Series D! Four years ago, founders Karan and Neel took a contrarian bet: the future of space isn't "smaller and lighter"—it’s high power and high mass. Today, consensus has caught up, and K2's vertically integrated platform is leading the market, as proven by their recent GRAVITAS mission success in orbit. Karan and Neel have built an extraordinary team defined by speed, discipline, and relentless engineering excellence. Thrilled to partner with them as they shape the future of space. CC: Manmeet Gujral Conner Lovely Seimi Park https://lnkd.in/gkVfASU5

Today, we're announcing that Whatnot has acquired Shaped, a machine learning company specializing in recommendation systems, search, and personalization. AI already powers discovery across Whatnot, helping buyers find the live shows, sellers, and products they'll love while helping sellers reach the right audience and grow their businesses. By bringing the Shaped team together with our existing AI organization, we're accelerating work already underway and deepening our investment in AI-powered discovery. We're excited to welcome the Shaped team to Whatnot and continue building the future of live commerce. https://lnkd.in/gMsjVJuB #ML #LiveShopping

Announcing AI Risk Management, a major milestone for DTEX and a new way to manage risk in the age of AI. 🤖 AI has crossed the trust boundary and is now embedded in how work gets done, often acting on its own across systems, data, and workflows. Most teams can see that activity. Few can tell if it actually creates risk. DTEX AI Risk Management (AIRM) closes that gap. Built on more than 20 years of behavioral intelligence and insider risk expertise, AIRM  helps organizations understand risk across people, AI agents, and the workflows that connect them by analyzing behavior and intent over time, not just isolated events. At the center of this are DTEX Agentic Defenders, a new class of AI-native agents that continuously hunts, investigates, and triages risk across human and AI-driven activity. As AI operates with greater speed and autonomy, security teams need defenses capable of keeping pace. AI Risk Management helps teams:  • Identify unknown AI tools and agents  • Understand intent, not just activity  • Prioritize real risk using proven behavioral models This is what makes DTEX different. AI risk is not just a visibility problem. Its about understand behavior and intent.   That's exactly how we help organizations manage AI risk and embrace the next era of AI with confidence. Read the blog to learn more: https://lnkd.in/ehGGXVXA

What we've been hearing from customers for years: The AI prototype works, the team is excited — and then it sits in a backlog for six months. Nobody can explain how it works. Risk won’t sign off. IT has no idea how to run it. We built Dataiku Cobuild for exactly that moment. Describe what you're trying to accomplish in plain language, and Cobuild builds out a complete AI project your whole team can see, understand, and approve before anything goes live: → Governed from day one: Inspectable AI projects, not prototypes. → Speed and control: Conversational building inside an enterprise-governed environment. → Closes the execution gap: From business intent to AI your teams can inspect, approve, and scale. The build-govern gap is real, and this is how you close it. See what Dataiku Cobuild can do: https://lnkd.in/eFZ7_7U6 #AISuccess #AIAgents

For most of software history, even if you understood a problem better than anyone, you still needed someone else to help you build the solution. That meant a lot of good ideas never became products. Today we’re sharing our first look at the global build economy: more than 50M projects built on Lovable and 720M monthly visits to Lovable-built projects. Ideas are becoming products, and products are turning into companies. The people building them are also different from what many would expect. They may not have a CS degree, venture capital, or a technical co-founder, but they have customer knowledge, industry context, and a clear sense of what needs to exist. It is still early, but the direction is clear: as this category matures, the impact will be greater than just more software. There will be more solutions, more companies, and more economic activity. Read the report: https://lnkd.in/dpydMSpQ