About
Partner at boldstart ventures investing at inception in the autonomous enterprise.
Articles by Ellen
Activity
11K followers
-
Ellen Chisa reposted thisEllen Chisa reposted thisMost robot demos are scripted. Generalist's GEN-1 is not. > GEN-1 was doing a task > Mid-task, an extra object was thrown into the bin > GEN-1 quickly adapted to the new environment > Still completed the task smoothly This is the difference between a scripted policy (seen in viral launch videos) and a real foundation model. GEN-1 was trained on 500,000 hours of proprietary, real-world dexterous manipulation data. It has NEVER seen that exact scenario before. This was the coolest demo at Automate Show
-
Ellen Chisa reposted thisEllen Chisa reposted thisCome see live demos of GEN-1 doing dexterous manipulation tasks at the Automate Show — the largest automation show in America, with our friends at Universal Robots and Flexiv Robotics If you’re at the show, stop by booths #1250 and #2844. Read more about GEN-1 and find the show floor map in the comments👇 For early access to our models, please reach out 📧 partnerships@generalistai.com #Automate2026 #Robotics #PhysicalAI #EmbodiedAI #Automation
-
Ellen Chisa reposted thisEllen Chisa reposted thisWe're proud to announce our investment in Generalist and their $400M round. Robotics is approaching the same inflection that large language models brought to software: the shift from narrow specialists to general-purpose foundation models that improve across every task at once. The crux of that shift is dexterity. It's the hardest unsolved problem in robotics, and the clearest path to everything that follows. Cracking it takes a rare combination of research depth and commercial focus, the kind that almost never coexists in one team. Generalist has both: a founding team among the most-published researchers in the field, relentlessly focused on what actually ships and compounds. We've followed their progress across successive generations of models, and the signal has been unmistakable: rapid adaptation to new robots and tasks, strong sample efficiency, and real commercial traction. Proud to be on the journey with Pete Florence, Andy Zeng, and the team. More on our conviction here, from Alex Kolicich and Vivek Gopalan: https://lnkd.in/gYSgzKrvAnnouncing our Investment in Generalist | Posts | 8VCAnnouncing our Investment in Generalist | Posts | 8VC
-
Ellen Chisa shared thisGeneralist just announced $400M in new funding - excited for what's to come - it's been amazing what has changed in the last two years (and even what's changed in the last few months since we originally shared a bit behind the scenes on our investment from boldstart ventures).Ellen Chisa shared thisToday, Generalist is announcing $400 million in new funding, bringing our total raised to more than half a billion dollars. This capital goes toward one mission: building physical AGI and making it useful to everyone. Millions of robots are operating in the world today. Billions more are coming — across factories, warehouses, laboratories, restaurants, farms, homes, and space. They will take many forms, but they will share one need: intelligence that can understand and act in the physical world. Last November, GEN-0 brought robotics into the pretraining era — the first proof that scaling laws apply to the physical world. This April, GEN-1 showed where that path leads: commercial viability, with 99% reliability on diverse tasks, execution up to 3x faster than prior SOTA, and emergent improvisational intelligence. These results are the compounding product of thousands of decisions made by a world-class team. And now we have more resources to keep pushing. The flywheel is beginning to turn. Better models create more capable physical systems. Real-world deployments generate the data that drives the next generation of models. This is how general intelligence will emerge in the physical world. We're proud to welcome new major investors including Radical Ventures (lead), 8VC, Union Square Ventures, Hanabi Capital, and Norwest, and new angels Bin Lin, Fei-Fei Li, and Naval Ravikant. Deep gratitude to all our existing investors, including NVIDIA's NVentures, Boldstart, Spark Capital, Bezos Expeditions, and NFDG, who continue to back this vision. We are still early. But the defining moments in AI come when research breakthroughs and product inflections compound on each other, and we're beginning to see exactly that. General intelligence will be born from the physical world. — The Generalist Team
-
Ellen Chisa reposted thisWe're hiring! If you're interested in software for hardware, come join us 🚀Ellen Chisa reposted thisWe're hiring a founding full stack engineer. If you want to work at a fast pace with the large manufacturers in the world solving the hardest problems, we're the place for you. Apply here: xnilio.com/jobs
-
Ellen Chisa reposted thisEllen Chisa reposted thisAI breaks differently. The tools haven't caught up. Today we're launching Kiln v1 to change that. Our 90-second video below shows what's broken about building AI ↓ The worst part: the team can't help. QA finds bugs they can't safely fix. A PM tweaks a prompt for one issue, and causes another. Designers can't guide the style or form. Everyone on the Kiln team spent the last decade at Apple and startups, shipping production AI to real users (read: before it was cool). From experience and conversations with other teams, we keep seeing AI teams get crushed between moving surfaces. 1) Leadership shifts product direction. 2) The underlying tech changes constantly. 3) QA finds bugs whose fixes regress something else. There's no good move. Stagnate on product, stagnate on tech, or stay constantly broken. We got tired of waiting for a better way to work. Before Kiln was even complete: 10,000+ users. 4,800+ GitHub stars. Engineers at top AI labs and product companies are already using it. Kiln is the workbench for the full AI development loop — evals, optimization, RAG, fine-tuning, agents, synthetic data. Highlights: → Automatic eval builder: build judges + eval datasets in ~10 minutes → Auto-Optimize: tunes prompts, models, tools, and subagents → Intuitive free app: an app everyone on the team can use. Mac, Windows, Linux. → Every major AI tool: RAG, tools/MCP, skills, reflective-optimization, and much more → Git-native collaboration: works for teammates who don't know Git → MIT-licensed Python library ships the same tasks to production And Kiln is built for the whole team: PMs produce evals instead of PRDs. QA gives feedback that lands in next-day optimizations. You can change product direction without breaking everything. AI engineers and data scientists get actionable evals, not nebulous specs and bug reports. Once your team works this way, the system starts improving itself. Kiln Assistant is our data-scientist agent, which reviews evals, designs experiments, and autonomously improves your quality, cost and latency. Curious: what does AI eval look like on your team today? Free. Local-first. MIT open-source library. Download today at https://kiln.tech -- or DM me to chat!
-
Ellen Chisa reposted thisEllen Chisa reposted thisMost GPU dashboards are lying to you. Your GPU utilization says 100%. It’s probably wrong. The standard GPU utilization metric doesn’t measure how hard your GPU is working, just that something is running. Popular tools like nvidia-smi, nvtop, rocm-smi, Weights & Biases, Amazon CloudWatch, Google Cloud Monitoring, and Azure Monitor all report utilization similarly. In reality, many systems are doing ~1–10% of the work they could, but the tools don’t tell you. Low utilization is costing more than you think. That’s exactly the problem we’ve been working on at Systalyze. We open-sourced #Utilyze to fix this. GitHub Repo: https://lnkd.in/eFGf-b9G Article: https://lnkd.in/e5v7kEHv #AI #Performance #Observability #MLOps
-
Ellen Chisa reposted thisExcited to be highlighted in the informal member feature! I'd like to thank everyone who has been a part of my journey, especially my wonderful coworkers at WHOOP, Nuro, and Lumia™, and mentors Whitney Q. Lohmeyer and Adam Broders. 📢 Shout out to the Sam Holland and Lindsey Gideon at informal and the awesome freelancers. 🚀 Lots of exciting stuff happening in 2026. Stay tuned! Check out the full interview below!Ellen Chisa reposted thisFrom the youngest EE at WHOOP to building a sub-1-gram earring that that monitors blood flow to the head, Utsav Gupta's career has taken him from biosensing wearables to autonomous delivery vehicles and back again, always following the same fundamentals: sensors, power, systems integration, and reliability. These days, he's exploring the Indian hardware ecosystem with a focus on deeptech, medical devices, and consumer products at early-stage startups. In our latest member feature, we sit down with Utsav to learn more about his background, how he got into the hardware ecosystem, what's firing him up these days, his professional superpower, and his advice to clients to new hardware.informal member profile: electrical engineer Utsav Guptainformal member profile: electrical engineer Utsav Gupta
-
Ellen Chisa reposted thisEllen Chisa reposted thisIntroducing Feeds. New realtime primitives for storing chat messages, workflow status, agent logs—anything that needs a live list. Plus, new APIs that allow AI to act as native users, showing live presence and editing sync engine documents. https://lblcks.io/DuqqCljIntroducing Feeds and APIs for Agent Workflows | Liveblocks blogIntroducing Feeds and APIs for Agent Workflows | Liveblocks blog
Ellen Chisa commented on a post
3w
https://boldstart.vc/news/generalistai-when-robots-start-to-improvise-welcome-to-boldstart/
-
Ellen Chisa liked thisSo excited the stealth period is now over! Excited for Robert and Gavin and the absolutely incredible team that they have built who buy into the mission so much they live next door to the Frankenstein office and testing lab that they have built. This team moves at lightning speed with everything they do along with executing at an intensity that is rare to find. The result is lead time in the inference category where they saw a need before many others when it came to new rack and chip solutions. Thrilled to have been able to partner with them over the past few rounds!Ellen Chisa liked thisWe're coming out of stealth. We've built our first racks after a successful A0 tapeout, $1B+ in customer contracts, and $800m raised. Early customer tests show us achieving SOTA throughput, latency, and power efficiency on inference workloads. Our first racks ship this summer. We're a team of 400+ engineers from NVIDIA, Google TPUs, Broadcom, SK Hynix, TSMC, & more. We're backed by Jane Street, HRT, Two Sigma, and Jump, with strategic investment from VentureTech Alliance. We're excited to deepen our partnership with the world's leading semiconductor manufacturer. Our Series B was led by Stripes, with participation from Ribbit Capital, Radical Ventures, Positive Sum, Primary, & Argo. Our inference systems are built to push the entire pareto curve on frontier models, including many-trillion parameter MoEs, long context, and agentic workloads. Today, we're sharing two breakthroughs to make this happen: Low-Voltage Inference (LVI) for high throughput workloads. Today, AI chips can't scale FLOPs without thermal throttling. As FLOPs utilization increases, AI chips draw more power and downregulate clock speed. This often results in sustained inference throughput under half of peak FLOPs. Chips in other industries solve the power problem by running at lower voltages. Bitcoin miners run at under 3x the voltage of AI chips! We’ve designed a new architecture to run our chip’s math blocks at under half the voltage of most AI chips. This enables multiple times the FLOPs density of AI chips today. Cluster-Scale Memory (CSM) for low latency workloads. Today's AI chips using HBM can’t achieve SRAM-level decode speeds due to memory subsystem and interconnect bottlenecks. SRAM-only chips have lower FLOPs density and memory capacity, sacrificing throughput. You’re forced to make a tradeoff: serve at much slower speeds, or run at low batch sizes and suffer from higher costs. When running large MoE models, token routing across experts requires sending data through a deep memory hierarchy and a networking switch to reach a destination expert. Each memory layer inherently adds latency; thus, the best layer is no layer. We’ve designed a new architecture that creates a shared low-latency memory pool across the entire scale-up domain. We use a proprietary ultra-low-latency, high-bandwidth interconnect to enable dramatically faster memory access across chips. Our HBM/SRAM hybrid design solves both memory capacity and mem2mem latency, enabling high throughput and interactivity simultaneously. CSM improves latency and avoids today's cost, reliability, yield, thermal, and compute tradeoffs of SRAM-only chips, 3D DRAM chips, or optics. We're scaling production as fast as possible. We built a 2MW datacenter in our office and opened a Taiwan factory for 24/7 engineering. Performance, roadmap, and more updates are coming this summer. If this excites you, join us to build the future of gigawatt-scale inference: https://lnkd.in/gaQBX-Fz.
-
Ellen Chisa liked thisEllen Chisa liked thisI’ve made the very difficult decision to move on from Microsoft. The past two and a half years have been inspiring, chaotic, intense, and deeply rewarding. I’ve had the opportunity to work on some of the most ubiquitous products in the history of software and reshape their futures with AI. I’ve been able to partner with world-class AI experts inside Microsoft AI, OpenAI, and Anthropic. I’ve been fortunate to work with some of the most collaborative, thoughtful, and kind teammates I’ve ever known. And I’ve been challenged in ways I never could have anticipated back in 2023. I’m sure I’ve learned and grown much more as a leader than I even realize right now in this moment. Thank you to everyone who made my time at Microsoft so memorable. I’m incredibly confident in you, and I’ll be rooting for you every single day. I’ll have more to share on what’s next for me very soon.
-
Ellen Chisa liked thisEllen Chisa liked thisA huge congratulations to Damien Lewke and the entire Nebulock team on announcing their $25M Series A led by David Waltcher at FirstMark! When I first met Damien he pushed my thinking on what the world of cybersecurity will look like in an AI first era. His view was that AI wouldn't just make security teams more efficient. It would also fundamentally change the tempo of cyber attacks. As models become more capable, attackers can move from initial access to meaningful impact in a fraction of the time. Security operations built around waiting for alerts and manually piecing together evidence simply weren't designed for that world. The insight that really resonated with me was what came next. In an agentic world, the limiting factor isn't intelligence—it's context. The systems that win will be the ones that can continuously build an understanding of the enterprise and deliver the right context, at the right time, to the right agent. Everything else becomes downstream of that. That vision has become Nebulock: a contextual security analytics platform that serves as the foundation for autonomous security operations. By continuously reasoning over enterprise context, its fleet of agents can proactively hunt for threats, analyze threat intelligence, secure Shadow AI, detect insider risk, and automate work that historically required large teams of analysts. If you have ever met Damien you know that he is relentlessly customer obsessed, endlessly curious, and one of the fastest executors I've worked with. In just nine months since coming out of stealth, Nebulock has deployed into Fortune 500 environments, built deep customer love, and assembled an incredible team including Suzannah Cooke (Snyk), Ian McShane (Arctic Wolf), Emily Dann (Palo Alto Networks), Christine Huynh (Sublime Security), Kevin Menezes (Tines), and many others. It's been a privilege to back Damien from the pre-seed alongside Rak Garg (Bain Capital Ventures (BCV), Apoorva Pandhi (Zetta Venture Partners), Will Lehmann (Step Function) Andrew Peterson & Nick Galbreath (Aviso Ventures). Congratulations to everyone at Nebulock and our friends at FirstMark! PS: Nebulock is hiring on all fronts from AI engineering, threat research and sales - Join.the.Hunt. Jon Sakoda, Alessio Fanelli Stacey Wueste Lauren Ipsen Ryan Sullivan Josh Kamdjou Andrew Morris Jeff Yoshimura Mario Götze Indu Sajeev Christopher Witter Charlie Moore Jon Oberheide Oliver Friedrichs Brian Gumbel Shayan Shafii Kevin Arsenault Lauren P. Patrick Gray
-
Ellen Chisa liked thisEllen Chisa liked thisAbout two years ago I met Damien when Nebulock was just an idea. Like every early founder meeting, we spent time talking about the market and the product. But what I kept thinking about after the meeting wasn't the idea. It was Damien. Every time we'd brainstorm an idea or challenge an assumption, I'd hear from him a few days later saying, "We tried it."He never seemed interested in debating something for weeks. He'd just go try it. I don't think that's a coincidence. Damien is an endurance athlete, and he builds companies the same way. He sets ambitious goals, shows up every day, and puts in the work. No shortcuts. One of the coolest parts of the last couple of years has been watching the team come together. Great people always have options, and many of them have chosen to build with Damien. That doesn't happen by accident. Seeing Nebulock announce its $25M Series A feels incredibly well deserved. It's also been a lot of fun partnering with Will, Dan, Rak and the rest of the syndicate along the way. Huge congratulations to Damien and the entire Nebulock team. Can't wait to see what comes next. 🚀
-
Ellen Chisa reacted on thisEllen Chisa reacted on thisMore than five years ago, I took a leap of faith and joined a stealth-mode food truck company in the suburbs of New Jersey, built on a wildly ambitious idea about what mealtime could become. It turned into the most challenging and rewarding chapter of my career. Today is my last day at Wonder. Over the years, I wore many hats across Growth and Operations, from incubating new functions and teams to taste-testing our way to the perfect delivery-optimized French fry. I had a front-row seat as we grew from those first food trucks to 130+ brick-and-mortars across the Northeast, making Wonder the fastest company in history to reach 100 locations, and betting bigger at every turn with acquisitions of Grubhub, Blue Apron, Spyce, Tastemade, and more. But none of that is what I'll remember most. What I'll carry with me are the people. To Marc Lore and the leadership team I've learned from: thank you for the trust, and for pushing me to be better than I thought I could be. And to every teammate across every chapter: you turned the most exhilarating years of my career into the best ones. I'll be cheering you on from the sidelines. These two photos bookend my five years at Wonder: me "driving" one of our original food trucks in New Jersey, and me in front of our first Massachusetts location, holding my newborn daughter. In between, Wonder found its footing, and I became a mom of two. More to come soon on what's next!
-
Ellen Chisa reacted on thisEllen Chisa reacted on thisHi folks! It's time for an update - I've been working behind the scenes and I’m ready to announce my Product and UX strategy consultancy, Forma Libera. I help companies build complex products like AI tools, data platforms, and developer tooling. The problems I solve don't look like product or UX problems at first glance. They look like a sales team hand-holding every account to get it to value, or a roadmap that's chasing the most recent customer complaints. I do the work to identify the root causes underneath these symptoms - which can be hard to see from inside your own company - and turn them into clear, evidence-backed next steps. What the solution looks like in practice is (fittingly!) a bit free form, and depends on what you're stuck on. It could be a roadmap, a more coherent structure for the product itself, or a team that comes out of it able to make these calls on its own. I'm taking on new work now, so if these kinds of problems sound familiar, let’s talk.
-
Ellen Chisa liked thisEllen Chisa liked thisA massive congratulations to my good friend Damien Lewke and the entire team at Nebulock who are today announcing their $25M Series A, led by David at FirstMark. If you know Damien, you know he truly cares about cybersecurity as a force for good, he is extremely competitive (in the best way), and loves running. Given that, this fundraising milestone felt inevitable. Two surprises for me, however, are that he didn't find a way to make the total $26.2M, and that David somehow avoided a 10+ mile run as part of the investment process. Nebulock is an AI-native, hunt-first contextual security platform. The 'hunt-first' part is what drew me in from our first meeting. The team has deep pedigree in threat hunting and built the platform around looking for threats rather than waiting on alerts. This proactive posture matters. Attackers are using AI to move faster and at greater scale; Nebulock empowers security teams to counter these threats before they become incidents. This has allowed them to earn the trust of security leaders at organizations ranging from the Fortune 500 to high-growth unicorn companies. That trust and traction catalyzed the funding round and the team is scaling across every function. They are indeed hiring, so please reach out if you are interested in joining a team shaping the future of cybersecurity operations. Inevitably, this is the part where I say I am proud to have supported Damien and the team since the beginning two years ago, alongside phenomenal co-investors such as Dan, Rak, Apoorva and many cybersecurity leaders. It has been quite the ride and I have loved every minute of it. Partnering with founders like Damien and companies like Nebulock is exactly why I love doing this job. We have co-hosted several CISO dinners, Security Speakeasy demo nights, and held impromptu strategy sessions in both our coworking space and on running paths around Boston. The photos attached here reflect many of these milestones, including the aftermath of the half-marathon Damien had me run with him to celebrate the Seed round in 2024. Thankfully we ran at my pace and not his. Congratulations to all of team Nebulock on this milestone. Company building is a hybrid of both a marathon and a sprint (that's the last run reference, I promise) and Nebulock is setting the pace (oops) for AI-native cybersecurity. For all they have already accomplished, I know the best parts of the journey are still ahead.
Experience & Education
-
boldstart ventures
View Ellen’s full experience
See their title, tenure and more.
or
Other similar profiles
Explore more posts
-
François Veron
Menlo Ventures just released its report “The State of AI in Healthcare”, and it strongly echoes what we felt when launching Newfund Heka. Among all verticals touched by AI, healthcare stands out. Because of data abundance and impact density: every incremental improvement can literally change lives. We’re seeing this first-hand with startups combining medical expertise and AI-first thinking to improve diagnostics, drug discovery, patient monitoring, and clinical workflows. The report highlights the same momentum: a sector still early, but already reshaping how care is delivered. At Heka, we've already invest in eight startups at the crossroads of AI and Brain Health. Stay tuned! https://lnkd.in/egvT2QUz cc Paul Herent, M.D., Msc Alexandre SANTINHO Jukka Laakso Adrian Trömel Anne-Sophie Saint-Martin
3 Comments -
Eurie Kim
For most of modern medical history, the “default human” in research was a 160 pound white man. Women were routinely excluded from clinical trials until the early 1990s. Even after policy changes, much of medicine — from cardiovascular research to sleep science — continued to rely on male-dominant data sets. The downstream effect? Gaps in understanding women’s physiology and health needs that we’re still working to close today. That’s why this announcement from Tom Hale and ŌURA matters so much. Today, ŌURA introduced its first proprietary large language model — purpose-built for health, and designed specifically to advance and empower women’s health insights. This isn’t a generic AI layer. It’s grounded in peer-reviewed clinical research and powered by the longitudinal biometric data Oura has been building for nearly a decade. Women’s bodies are dynamic — hormonal cycles influence temperature, sleep, recovery, cardiovascular strain. Personalized health requires context and pattern recognition over time, not one-size-fits-all advice. And just as important: privacy. As AI becomes a first stop for health questions, we Forerunner believe trust IS the product. Building a private, secure model reflects a commitment to responsible AI and long-term stewardship of deeply personal data. I’m especially proud that Oura is partnering with webAI — another Forerunner portfolio company I’m fortunate to be involved with — on this ground breaking initiative. Bringing together biometric intelligence, secure AI infrastructure, and a mission to advance women’s health is exactly the kind of ecosystem thinking that moves markets forward. As an investor and board member at Oura, I couldn’t be more inspired about this meaningful step forward towards what the future of healthcare can and should look be for us all. Congratulations to Tom Hale, Dorothy Kilroy, Ricky Bloomfield, Holly Shelton, Shyamal Patel and the entire ŌURA team. Onward and upwards!
9 Comments -
Ken Nelson
Katie Jacobs Stanton , founder of Moxxie Ventures, shared some details of how she and her team evaluate healthcare AI startups. She stressed the need for founders with deep industry expertise, differentiated products and realistic distribution strategies. MedCity News MedTech Innovator LSI American Heart Association Heart Rhythm Society HRX HeartX Accelerator MedAxiom - An ACC Company American College of Cardiology MedTech World MedTech Strategist
3 Comments -
Andrew J Scott
Have been excited to back Ross and his team since the start of the Limbic journey; today is an important next step in the evolution of Limbic as the digital plumbing for the mental health world: "Starting today, the Limbic Layer is available as an API to responsible teams building in mental health AI"
1 Comment
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content