Foundation Capital @FoundationCap

Making today’s insurance processes faster is only part of the AI opportunity. Fulcrum co-founder and CEO Arjun Mangla argues that today’s impressive automation gains could become tomorrow’s baseline. The bigger opportunity is to rethink how insurance works. Take a business adding five trucks to its fleet. Today, updating its coverage means emails, forms, and repeated follow-ups between the business, its broker, and its insurer. Arjun envisions insurance connected to a company’s business systems, so coverage can adjust automatically as the business changes. Hear more about what that future could look like on AI in the Real World, out now: → YouTube: youtu.be/wJMMK-msDBY → Spotify: open.spotify.com/episode/5eAOZi… → Apple Podcasts: podcasts.apple.com/us/podcast/ins…

Some entrepreneurs are born. Arjun Mangla, co-founder and CEO of Fulcrum, shares how his father’s passion for building a company gave him the foundation to start his own. Stay tuned for our latest AI in the Real World episode, coming soon.

We're excited to have led IPID's $16M Series A! Today, payments have gotten much faster, but banks still don’t always have the information they need before money moves. IPID enables companies to verify who they're paying, whether or not the account is open, and what the account's currency is before any money is sent. The company already reaches more than 4B accounts, covering 85% of the world's banked population, and they're continuing to grow. We're thrilled to partner with Damien Dugauquier and the IPID team.

None of our security systems today were built with agents in mind. In our latest B2BaCEO podcast, @SkyflowAPI co-founder and CEO @anshublog joins @ashugarg to talk about what he’s learned securing sensitive customer data across data stores, models, and agents. It aligns with a problem he’s been chasing for over two decades. “At some point you realize, you know what? There's probably a $100B company to be built just protecting the most sensitive data for customers at all these companies,” he shares. Tune in to the full episode: Spotify: open.spotify.com/episode/04kudr… Apple Podcasts: podcasts.apple.com/us/podcast/how… YouTube: youtu.be/_OwWEl8i31U

We've raised $4.9M to build @tndminc - The AI system for hardware engineering. Engineering ambition has outgrown the systems built to carry it. Tandem connects requirements, design intent, CAD changes, reviews, and validation. The next great machine is waiting on its system.

The best way to understand new technology? Test it out for yourself. @JayaGup10 is letting AI agents manage her life operations. The day Muse launched, she used it to successfully find and apply for a new apartment and kick off a bidding process for a parking spot in San Francisco. Through hands-on learning, she's building insight about what the next great consumer AI agent will look like. For example, in her experiments with Instinct, the agent changed her passwords but didn't tell her. Communicating everything to the user—and being thoughtful about security and privacy concerns—will be key for adoption. Watch the full interview with @theinformation: youtube.com/live/5j1qi_4yn…

@Maximor_ai is rearchitecting finance from the ground up. Their co-founder and CTO, @ajkrish95, joined @JayaGup10 to unpack how they think about the models that power their automated finance platform. Full episode of AI in the Real World dropping soon.

meet @mostik_ai! what happens when you put 12 PhDs in one room for four months? first place on the ARC-AGI leaderboard, which I can't say much about while the competition is still running. and this, which I can. everyone's arguing about whether open models will catch up to frontier models. we think it's the wrong question. here's the one we pose: why does a frontier model have to generate your answer at all, when the only thing you need from it is the reasoning? we do this by enabling models to communicate in latent space. through our protocol, hidden states pass straight from a frontier model into a small one running on your infrastructure -- no text between them, and neither model is fine-tuned. two models from different families, sharing reasoning, both left untouched. how do we know it works? we tested it on a setup where a 753B model reads the problem, and a 4B edge-class model writes the answer. with this approach, we get results 80% as accurate as the frontier model, but at 20x faster performance. we're committed to preventing frontier model lock-in and are already partnering with inference providers to accelerate open-weight adoption. we've done this between 15 of us, in four months, 12 PhDs and a Fields medalist, backed by @generalcatalyst WIRED has the first external account of the company and the work: wired.com/story/russian-… full writeup, the setup, and all the numbers: mostik.ai/read-more

What does it take to run a billion-dollar brand with just a handful of employees? Today, Seel runs over $6B in GMV with fewer than 10 support reps and a 99%+ resolution rate. Zack Peng, co-founder and CEO, discusses with @fifaifofo what he’s learned from using AI to automate the operational work behind e-commerce. Check out the latest episode of AI in the Real World, out now wherever you listen to podcasts. → Spotify: open.spotify.com/episode/61FR8G… → Apple Podcasts: podcasts.apple.com/us/podcast/ai-… → YouTube: youtube.com/watch?v=zCEw9y…

We scored 20 leading VC websites on how readable they are to an AI agent. The median score? Just 69/100. Only two sites had a valid llms.txt. Most software (not just websites) was built for humans, but agents are about to become its heaviest users. In our latest article, @joannezchen and Leo break down which parts of the stack are under the most pressure to change, and where we think the biggest opportunities are: execution environments, shared artifact layers, and business process systems. Read it here and check out the full leaderboard: foundationcapital.com/ideas/rebuildi…

Introducing MachineView. It scores how readable your website is to an AI agent, from 0 to 100 (for free). @realleolu and I ran 40 leading websites through it. Agents are becoming the web's new consumers and most websites aren't ready. Here’s how it works:

We're proud to share that Foundation Capital made @TIME's list of top venture capital firms this year! We're in the top 25: time.com/article/2026/0…

Congratulations to @Jason_Lopatecki, @AparnaDhinak, and the whole @ArizeAI team on their $915M acquisition by @Dynatrace! 🎉 We led Arize’s seed in 2020, when AI observability was barely a category. Today, Arize runs inside companies from Uber, Instacart, PagerDuty, Atlassian, and Reddit. We’re proud to have partnered with Jason and Aparna; this is the second of Jason’s companies we’ve backed. Read more about the journey in @ashugarg's story on our blog: foundationcapital.com/ideas/making-a…

Arize began with the mission to make the world’s AI work. Today, we’re excited to share that we have signed a definitive agreement for Dynatrace to acquire Arize to help accelerate our mission. The two worlds of AI and software observability are coming together to create this next chapter. Arize exists because of the trust of our users, contributors, and community. Thank you for trusting us with your most important AI systems and building this category with us. You make our products better everyday. Read more from Jason in: arize.com/blog/a-new-cha…

Jason and I started Arize 6+ years ago with a simple proposition that headlined our seed deck: “We Make the World's AI Work” Today we are announcing we’ve entered into a definitive agreement to be acquired by Dynatrace to accelerate that vision. We made a bet years ago that the explosion of AI would require bespoke infra tools. We knew AI systems weren’t going to behave like traditional software because we were building the next generation of intelligence. That bet became the market's first AI Observability platform. Fast forward to 2026. The world has changed dramatically. Every company in the world is scaling up its use of agents. AI is everywhere and agents are reshaping business after business. What wasn’t clear years ago and is crystal clear today is that agent systems and software are deeply connected together in ways that none of us fathomed. - The best general agents are coding agents. - Skills and tools for agents are mixtures of prompts and code. - Prompts that control agents live in code repos. - The logs and traces used to debug your software also help you debug your AI systems. - The path to AGI is likely through coding agents. The two worlds of AI and software observability are coming together to create the future of observability. Lots more to share in the coming days, but I wanted to thank folks who made this possible: To the Arize team - startups are about the people you build with, and I’ve never met a better group of humble rockstars. The markets changed, our product evolved, but the camaraderie of the team through this journey is what I’m really proud of. To Jason, being in the trenches with you for the past 6 years has been the journey of a lifetime. You’re a visionary, and the best entrepreneur I know. Everything I know about entrepreneurship comes from watching you execute day in and day out - you always figured it out. Arize wouldn’t be Arize without you. To our customers, thank you for trusting us and building this category with us. You made the product better everyday, and inspired us with the products you’re building. To the open source community around Phoenix and OpenInference, developer trust and open standards are the foundation this category gets built on. And to our investors, advisors, and early believers who backed us when AI observability was not a budget line or even a real category yet, thank you. Onwards to the next chapter! arize.com/blog/a-new-cha…

"The misconception is that AI is just about job automation or job loss." Foundation Capital’s Joanne Chen says AI will reorganize companies and remove bottlenecks, potentially allowing smaller teams to build much bigger businesses bit.ly/4wRbOry

“If AI means more automation, doesn't that mean fewer jobs?" @EdLudlow asked me this on @BloombergTV and @technology today. Here's what I told him: I don’t think so. Here's why. At @FoundationCap we're seeing AI remove bottlenecks in organizations. And yes, removing bottlenecks means you can start with fewer people. But if you're an ambitious founder and you realize you can build a billion-dollar company with 100 people, why stop there? Why not hire 500 and build something much larger? Ambition doesn't shrink when costs do. When the cost of intelligence drops, the size of what founders can build goes up. The misconception is that this is just a story about job automation. I think what's actually going to happen is AI will create a whole host of use cases we haven't thought of yet. Think about the early days of mobile. The first app I ever used was a beer-drinking app - you tilted your iPhone and the beer poured toward you. Five years later, Uber came along and changed how we move through the world. I believe we're going to see the same thing with AI. How organizations actually look in 10 years is hard to say. But I don't think they’ll look like they do today - and I don't think the story is fewer people. Thanks again for having me on @EdLudlow. Great conversation.

📣 @JayaGup10, Partner at @FoundationCap, at Agent Night.

x.com/i/article/2083…