Chemistry

Comfy MCP is now local and open-source! When we shipped Cloud MCP in June, the response was immediate and consistent: make it work locally. So we did and it's fully open source. Connect Claude, Codex, Cursor, or any MCP client to your local ComfyUI. Your agent reads the GPU you actually have and gives you a straight answer on whether a model is worth running before you commit to the download. It reads every node and model you've installed. It handles the setup that usually stops people at step one. It is now the easiest way to help with your local Minimax H3 workflows! Cloud MCP still does everything it did. Tell your agent where a job goes, or let it decide. Details: https://lnkd.in/gXF_m5Fm

Lots of the airtime around AI goes to the Bay Area schools like Stanford and Cal, but Brown University alums are building at many of the best AI companies. I had a chance to bring a group of them together last week—folks from OpenAI, Anthropic, Cursor, Notion, Thinking Machines Lab, Fireworks AI, Together AI Glean, Figma, a few stealth founders, and more—and was blown away by this community. We're planning to make these gatherings a regular thing. If you're a Brown student or alum working in AI and would like to join future events, send me a DM and I'll keep you in the loop. Big thanks to the Chemistry team—Kristina Shen, Ethan Kurzweil, Bohan Lou, Nick Faulkner—and Lauren Kolodny for helping make it happen.

I spent last month in China meeting most of the tier-1 investors and the management teams at many of the major robotics and biotech companies. The most surprising thing: Chinese founders routinely sign term sheets that make them personally liable if the company doesn't return capital within 6–8 years. Not the company. The founder. That single term explains most of the rest. Why every founder I met was sprinting for an IPO next year — not because conditions are ripe, but because China has essentially no M&A market. Nobody acquires. So you list, or you owe. It also explains the layer I didn't expect at all: FAs. Not wealth managers, but early-stage bankers who package and broker rounds, take 2–5%, and serve as the founder's first point of contact instead of the VC. Chinese firms have outsourced sourcing and first-pass diligence to a middleman. Why does that layer exist? Because China has no cold outreach. No cold email, no cold DMs. LinkedIn never took hold and neither did the domestic copycats. Everything runs through 关系. FAs exist to manufacture relationship liquidity in a market that produces none on its own. The conclusion I keep coming back to: harsher capital markets produce leaner, more ruthless companies that out-execute globally — and sometimes more fraud, because the pressure on founders is existential rather than reputational. Everyone in the Valley has a take on whether China is winning AI, robotics, and biotech. Almost nobody understands how the capital works. More content on X @loubohan Full notes below. 👇

When hiring executives, Travis Kalanick prioritizes one skill above all others: the ability to solve hard problems. "There's two things you want an executive to do. You want them to be able to organize, manage, and lead at scale. You also want them to be epic and strategic problem solvers... the best executives are the ones that are doing both well." "I have come to the conclusion over my years, that problem solving is the most important thing. If you get somebody who organizes and manages well but cannot solve a problem, they're going to be doing ridiculous stuff in a super organized way. The only constraint on your imagination is management capacity. But what is management capacity? It's really problem solving at scale."

Today, we're announcing a $1.7B equity investment in Atoms, led by Andreessen Horowitz. As part of this round, our operating businesses have merged into one Atoms equity structure. At Atoms, we’re on a mission to build physical automation to transform industry and move the world. If you look around, everything is grown or mined, manufactured and moved. That is industry - the large-scale, organized production of the things civilization runs on. Mining, construction, heavy transport, food production. Trillion-dollar sectors that have gone barely a generation without a genuine technological transformation. That is about to change. The impetus is Industrial AI - systems of software, sensors, robotics, and AI that automate the operations of entire industrial sectors. Industrial AI is what will power the next industrial revolution. Atoms is the OEM for atoms-based computers across the industries that move the physical world: - Atoms Food - infrastructure for better food. - Atoms Mining - more productive mines to power earth’s industries. - Atoms Transport - wheelbase for robots. For eight years, we've been building quietly. This round is fuel for the next phase - more machines, more deployments, more of the hardest problems in the physical world attacked at once. This is going to take builders with vision, grit, systems thinking, resilience, and a fighting spirit. Nature is going to throw everything it has at the builders in this new industrial age and it’s going to take humanity’s strongest to stay the course and get these complex systems and industries over the finish line. Thank you to Andreessen Horowitz and our capital partners for backing the progress machines we build to power this next industrial revolution. If you’re looking for a challenge, come join the team. Link in comments. -- Bain Capital Ventures (BCV), Uber, Fifth Wall, A*, Chemistry, K5 Global, Abstract, SV Angel, Alpha Square Group

I had a front-row seat to Travis Kalanick building Uber, and I learned what a generational founder looks like. Now he, Anthony Levandowski, and Gautam Gupta are going after industrial AI across food, transport, and mining — three multitrillion-dollar industries ready for transformation. Chemistry is proud to be part of Atoms' $1.7B round led by Andreessen Horowitz — the largest check in our firm's history, and the largest I've personally written in my 13+ years in venture. The Age of Atoms starts now. The robots are reporting for work. 🤖

Today we raised a $120M Series C at a $1.2B valuation, led by Menlo Ventures. We now have over $220M to become the standard for how healthcare providers transform the patient journey with AI. As a medical student at UCSF, I watched patients fall through the cracks for reasons that had nothing to do with medicine. Healthcare has spent decades improving clinical outcomes, yet patients were still repeating their story, navigating disconnected systems, and too often giving up before they got the care they needed. I couldn't accept that patients spent more time on hold than with their doctors. It felt like one of the biggest unsolved problems in healthcare. So I left to build Assort Health. Our first chapter was learning how complex each practice's care protocols really are. It took us months to build a single agent. That's why we built Synapse, our agentic model. It gets sharper with every edge case it sees, and with 190M+ patient interactions behind it, no one's model has seen more. That's what lets non-technical people build agents in weeks, even for the most complex workflows, with industry-leading automation rates. This is how we grew revenue 20x in 15 months, as leading providers chose the one platform proven to handle the complexity of healthcare workflows at scale. But this was always meant to be just the first chapter. The first interaction isn't only where care begins. It's where patient context is created. Get that right, and every step after becomes easier. Today we're a platform for the whole journey. Our AI agents answer calls, schedule appointments, manage referrals, complete intake, refill medications, reach patients proactively, and collect payments, all personalized to each patient. Grateful to Jean-Paul (J.P.) Sanday, Matt Murphy and the team at Menlo Ventures for leading, and to Lightspeed, Felicis, First Round Capital, Chemistry, Quiet Capital, and Joe Montana for backing us again. To the Assort team and my co-founder Jeffery L.: none of this happens without your relentless focus and care. Fired up to build Healthcare that Remembers alongside you. Still Day One.