Madrona

Looking forward to speaking with Pablo Casilimas at #eMergeAmericas 2025, where #tech leaders and innovators come together to shape the future of technology! Join me in #Miami on March 27 as we explore the evolving landscape of venture capital, the strategies behind successful investments, and what it takes to build enduring companies in today’s market, and more. Can’t wait to connect with you all!??? ? #globaltech #floridatech #Miamitech #innovation

Thinking about launching a B2B startup? Don’t skip this step. A lesson Edward Wu, founder of Dropzone AI, learned before starting his own company that he thinks others should follow: Working at an established B2B company first can give you an edge. ?? Why? Because understanding what great engineering, sales, marketing, and customer success look like before you start your own company will save you years of trial and error. This is just one of the insights Edward shares in our latest episode of Founded & Funded with Madrona Partner Vivek Ramaswami, where he breaks down: ? Why he left academia to build in the real world ? The ‘aha’ moment that led him to start Dropzone ? The most competitive market in cybersecurity—and how he’s standing out ? How AI is tackling one of the biggest security challenges today Watch the full conversation here: https://lnkd.in/gF9ck3sR

If 70% of your revenue comes from just 20% of your reps, you don’t have a scalable sales engine — you have a hero-driven model that’s bound to break. Sales success isn’t just about closing deals — it’s about building a predictable, repeatable revenue engine that doesn’t rely on last-minute heroics. In a recent GTM AMA with portfolio CEOs, Anna Baird and Loren Alhadeff explored what it takes to scale revenue the right way — and we're sharing their insights from that discussion with you all. Some of the topics discussed include: 1) Why pipeline is more than a sales metric 2) The importance of a pipeline council for cross-functional accountability 3) Focusing on rep attainment vs. quota achievement 4) Aligning compensation plans with business objectives Read their full playbook here: https://lnkd.in/gxzZw8NY

Is it still a risk-on environment? Are AI-native companies poised to outpace incumbents? And just how efficient can these small, high-growth teams get? These were just a few of the big questions we tackled at?The Montgomery Summit?last week. Grateful to have shared the stage with?Monica Adractas, George Mathew, and Jamie Montgomery?— engaging in a thoughtful discussion (and a bit of a friendly debate!) on where capital, AI, and markets are headed. A few takeaways from our discussion: ???We are still in a risk-on environment?— not just because of macro stability, but because of capital?waiting to be deployed. Private equity alone has?$1T+ on the sidelines,?and strategics are back in the game. ???The AI-native winners are starting to emerge.?For years, the focus has been on the?foundation models and tooling layers?— but now, we’re seeing?real breakout application-layer companies.?The next?Salesforce or ServiceNow?could be built in the next 24 months. ???The “Small Team Meme” is real.?Companies are scaling revenue?faster and more efficiently?than ever before. Some startups are growing to?$50M+ in ARR with teams of fewer than 50 people.?But how sustainable is this??Where’s the limit? If you were not able to make it to The Montgomery Summit or this panel, and you’re thinking about?funding, growth, or AI strategy?— this conversation is well worth your time. Watch the full discussion here: https://lnkd.in/gNPSxJcV

?? AI-powered SOC Analysts are here—and they’re changing security. ?? Security teams are drowning in alerts. The industry’s biggest challenge isn’t detecting threats—it’s acting on them fast enough. What if AI could do more than just detect? What if it could think and respond like a security analyst? ? That question is what led Edward Wu to found Dropzone AI after spending nearly a decade at ExtraHop where he helped pioneer AI-driven security detection. Today Dropzone is one of the leading AI SOC Analysts, handling Tier 1 alert triages and autonomously investigating every alert with no prompt required. ? In this episode of Founded & Funded, Edward and I dive into: ?? The "aha moment" that led Edward to leave a successful career to start Dropzone ?? Why most AI security startups will fail, and what it takes to stand out ?? How Dropzone landed UiPath as a customer — without a single cold email ?? The future of security in an AI-first world ? Watch or listen to the full conversation for insights from Edward on what it takes to build in one of the most competitive categories in security?today. ? YouTube:?https://lnkd.in/gxZBbrQW Full Transcript:?https://lnkd.in/gFN_ghtU

Most revenue teams live or die by a handful of top performers. But if 70% of your revenue comes from just 20% of your reps, your sales engine isn’t built to scale. In a recent Madrona GTM AMA, we broke down what it takes to move beyond a quota-chasing culture and build a predictable, repeatable revenue machine. Plus, in Aspiring for Intelligence, our investors go hands-on with AI agents — how easy (or not) is it to hack one together as a non-technical user? The New York Times took its readers into the 'vibecoding' future, with Bolt (by StackBlitz, a Madrona portfolio company), showing all that's possible when AI tools let anyone build software just by describing what they want. And with 11 Madrona portfolio companies and 2 portfolio CEOs up for GeekWire Awards, we’re cheering them on and encouraging the community to vote. Anna Baird Loren Alhadeff Jon Turow Vivek Ramaswami Sabrina Wu

Replit CEO Amjad Masad showed us a glimpse of the future back in December. His doctor gave him a PDF sleep diary to fill out. Instead of struggling with the clunky form, he used an AI agent to convert it into an interactive app with CSV export in minutes. Now, Amjad has the technical skills to code this app traditionally – but probably wouldn't have invested the time. When AI-powered tools let us create apps in minutes instead of months, solutions become practical for problems that could never justify traditional development. Each of us lives with countless small inefficiencies – paper cuts we've become blind to. These workarounds persist because custom software wasn't worth the effort. But these paper cuts add up. Vibecoding changes the equation. The shift from software created by the few to software created by the many will transform how we work, create, and solve problems together. More here: https://lnkd.in/eAykpBNB

Ginormous models add friction and limit model adoption. Throwback to the "Navigating the Model Landscape" panel at Madrona's #IASummit in October, where Ai2 CEO Ali Farhadi made a key point: If a model is too massive to fine-tune, too complex to serve, and doesn’t fit within the constraints of existing infrastructure, it won’t see widespread adoption. This discussion with leaders from Microsoft (Asha Sharma), Cohere (Saurabh Baji), and Newcomer (Eric Newcomer) covers the competitive dynamics of foundation models, trends in model development, and how enterprises are thinking about AI adoption today. Watch the full conversation here: https://bit.ly/41Lwy6W

What was the game-changing moment for MosaicML that helped them gain massive traction across industries? Madrona Partner Jon Turow recently sat down with Databricks Chief AI Scientist Jonathan Frankle to dig into?MosaicML’s journey — from an early bet on efficient model training to a defining moment when they released the MPT models. That milestone didn’t just showcase technical capability; it put them on the map and ultimately helped pave the way for their acquisition by Databricks. In this conversation, Frankle shares key lessons from that experience, including how AI startups can balance visibility and execution, why enterprise AI adoption follows a hype cycle, and how Databricks is now expanding its mission to democratize AI by integrating it more deeply with data. Catch their full conversation on YouTube: https://lnkd.in/gQEibQp3