About
I’m a product architect focused on designing and scaling AI-first products — from 0→1…
Articles by Anand
-
The Anatomy of Highly Resourceful Product Managers: Breaking Myths and Defining Excellence
The Anatomy of Highly Resourceful Product Managers: Breaking Myths and Defining Excellence
In the world of product management, certain individuals stand out as forces of nature. They operate with a unique blend…
59 Comments
Activity
51K followers
-
Anand Arivukkarasu reposted thisAnand Arivukkarasu reposted thisSXSW 2026 Panel: AI for SaaS: Huge thank you to Ilmo Lounasmaa and Softlandia for hosting, and to Hugh Forrest, Anand Arivukkarasu, Liz Bacelar, Joshua Liberson, and Markus Hoefinger for one of the most honest AI conversations I have heard in a long time. The tools can build almost anything now. So why are so many teams still building the wrong things? Anand called AI models "drunken geniuses." Pretty drunk. But surprisingly capable if you slow them down, give them a role, and make them think carefully. The problem? They are also deeply agreeable. They will tell you every idea you have is a billion-dollar company. He joked about buying a hundred domains on GoDaddy because of this. That is funny. It is also a real warning. I attended the AI for SaaS panel at SXSW 2026 and walked out with a notebook full of things worth thinking harder about. As a Product Manager, the thread I could not put down: almost every AI dollar right now is going into building things faster. Almost nothing is going into figuring out what to build and why. That navigation problem is still entirely human work. And it just became the most important work in the room.SXSW AI for SaaS: The Drunken Geniuses Are Here. Now What? Reflections from a Panel at SXSW 2026SXSW AI for SaaS: The Drunken Geniuses Are Here. Now What? Reflections from a Panel at SXSW 2026
-
Anand Arivukkarasu shared this𝗪𝗲'𝗿𝗲 𝗴𝗶𝘃𝗶𝗻𝗴 𝗮𝘄𝗮𝘆 𝗮 𝗰𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗔𝗜-𝗽𝗼𝘄𝗲𝗿𝗲𝗱 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗢𝗽𝗲𝗿𝗮𝘁𝗶𝗻𝗴 𝗦𝘆𝘀𝘁𝗲𝗺. One .md file. Attach it to Claude Cowork, ChatGPT, or Gemini Antigravity. It covers the 4 areas of PM skills: There are going to be two types of PMs. The ones who use AI as a real operating layer — for discovery, strategy, execution, and communication. And the ones still copy-pasting frameworks from blog posts. 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗦𝗲𝗻𝘀𝗲 — discovery, customer insight, problem framing, prioritization, roadmapping 𝗘𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻 & 𝗔𝗻𝗮𝗹𝘆𝘁𝗶𝗰𝘀 — PRDs, launch planning, decision docs, success metrics, retention and activation analysis, post-launch review 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆 — positioning, competitive advantage, where to play and how to win, long-term roadmap thinking 𝗟𝗲𝗮𝗱𝗲𝗿𝘀𝗵𝗶𝗽 & 𝗖𝗼𝗺𝗺𝘂𝗻𝗶𝗰𝗮𝘁𝗶𝗼𝗻 — stakeholder management, exec updates, cross-functional alignment, structured decision communication Add your product context or product folder— your company, competitors, OKRs, stakeholders, writing style — and it becomes a PM copilot that actually sounds like it belongs on your team. 𝗪𝗵𝗮𝘁'𝘀 𝗶𝗻𝘀𝗶𝗱𝗲: → 𝟳𝟬+ 𝗳𝗿𝗮𝗺𝗲𝘄𝗼𝗿𝗸𝘀 → 𝟳 𝗿𝗲𝘃𝗶𝗲𝘄 𝗮𝗴𝗲𝗻𝘁𝘀 𝘁𝗵𝗮𝘁 𝘀𝗶𝗺𝘂𝗹𝗮𝘁𝗲 𝗮 𝗿𝗲𝗮𝗹 𝗰𝗿𝗼𝘀𝘀-𝗳𝘂𝗻𝗰𝘁𝗶𝗼𝗻𝗮𝗹 𝗽𝗮𝗻𝗲𝗹 → 𝗧𝗲𝗺𝗽𝗹𝗮𝘁𝗲𝘀 𝗳𝗼𝗿 𝗣𝗥𝗗𝘀, 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻 𝗱𝗼𝗰𝘀, 𝗹𝗮𝘂𝗻𝗰𝗵𝗲𝘀, 𝗮𝗻𝗱 𝘄𝗲𝗲𝗸𝗹𝘆 𝗿𝗲𝘃𝗶𝗲𝘄𝘀 → 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄𝘀 𝘁𝗵𝗮𝘁 𝗴𝗲𝘁 𝘀𝗵𝗮𝗿𝗽𝗲𝗿 𝘁𝗵𝗲 𝗺𝗼𝗿𝗲 𝗰𝗼𝗻𝘁𝗲𝘅𝘁 𝘆𝗼𝘂 𝗴𝗶𝘃𝗲 𝗶𝘁 We run ours on Claude Cowork — it reads the OS, connects to Slack, Notion, Google Drive, and runs product workflows end to end. Not just answering questions. Actually doing the work. This isn't a prompt pack. It's an operating system for product work. 𝗩𝟭 𝗶𝘀 𝗳𝗿𝗲𝗲 — first 100 only, first served.. 𝗣𝗹𝗲𝗮𝘀𝗲 𝘀𝗵𝗮𝗿𝗲 𝘁𝗵𝗲 𝗽𝗼𝘀𝘁, 𝗗𝗠 𝘂𝘀 𝗮𝗻𝗱 𝗖𝗼𝗺𝗺𝗲𝗻𝘁 "𝗣𝗠 𝗢𝗦" 𝗮𝗻𝗱 𝘄𝗲'𝗹𝗹 𝘀𝗵𝗮𝗿𝗲 𝘁𝗵𝗲 𝗚𝗼𝗼𝗴���𝗲 𝗱𝗿𝗶𝘃𝗲 𝗰𝗼𝗻𝘁𝗲𝗻𝘁. V2 is already in progress and coming soon — Tech architecture, system design, agentic workflows and architecture that run multi-step product processes end to end, automated review pipelines, and proprietary frameworks like 𝗧𝗨𝗙𝗙𝗪𝗔𝗥 𝗮𝗻𝗱 𝗡𝗠𝗕𝗔 that go deeper than anything in the public domain. V1 gives you the foundation playbook. V2 runs the plays.
-
Anand Arivukkarasu shared this𝗔𝗻 𝗲𝗮𝘀𝘆 𝘄𝗮𝘆 𝘁𝗼 𝗹𝗲𝗮𝗿𝗻 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗔𝗜 — 𝗣𝗮𝗿𝘁 𝟭 𝗼𝗳 𝟰 A lot of my learning on Agentic AI has come from 𝗵𝗲𝗹𝗽𝗶𝗻𝗴 𝗺𝘂𝗹𝘁𝗶𝗽𝗹𝗲 𝗔𝗜 𝘀𝘁𝗮𝗿𝘁𝘂𝗽𝘀 𝗯𝘂𝗶𝗹𝗱 𝗿𝗲𝗮𝗹 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝘀 in our studio. Across teams and use cases, I kept seeing the same gap: People understand LLMs at a high level, but they don’t always have a clear 𝘀𝘆𝘀𝘁𝗲𝗺-𝗹𝗲𝘃𝗲𝗹 𝗺𝗲𝗻𝘁𝗮𝗹 𝗺𝗼𝗱𝗲𝗹 for how agentic products actually work — especially when memory, planning, tools, control, evaluation, and recovery all come together. So I put together a 𝘃𝗶𝘀𝘂𝗮𝗹 𝗿𝗲𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝗴𝘂𝗶𝗱𝗲 to make this easier to learn and easier to build. I’m sharing it as a 𝟰-𝗽𝗮𝗿𝘁 𝘀𝗲𝗿𝗶𝗲𝘀. ### Part 1 covers the foundations 𝘞𝘩𝘢𝘵 𝘮𝘢𝘬𝘦𝘴 𝘢 𝘴𝘺𝘴𝘵𝘦𝘮 𝘢𝘯 *agent** (vs a chatbot) 𝘛𝘩𝘦 𝘤𝘰𝘳𝘦 *agent loop** (perceive → think/plan → act → learn) * 𝗠𝗲𝗺𝗼𝗿𝘆 (short-term vs long-term) * 𝗣𝗹𝗮𝗻𝗻𝗶𝗻𝗴 and decision flow * 𝗧𝗼𝗼𝗹𝘀/𝗮𝗰𝘁𝗶𝗼𝗻𝘀 and system interaction * 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 𝗮𝗻𝗱 𝘀𝗮𝗳𝗲𝘁𝘆 * 𝗘𝘃𝗮𝗹𝘂𝗮𝘁𝗶𝗼𝗻 * 𝗧𝗲𝘀𝘁𝗶𝗻𝗴 𝗮𝗻𝗱 𝘀𝗶𝗺𝘂𝗹𝗮𝘁𝗶𝗼𝗻 * 𝗗𝗲𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁, 𝗺𝗼𝗻𝗶𝘁𝗼𝗿𝗶𝗻𝗴, 𝗮𝗻𝗱 𝗿𝗲𝗰𝗼𝘃𝗲𝗿𝘆 𝘊𝘰𝘮𝘮𝘰𝘯 *failure modes** in agent systems ### Who this is for * Founders building AI products * PMs shaping agentic workflows * Engineers moving from demos to production * Teams that want a cleaner model of how AI systems work 𝗣𝗮𝗿𝘁 𝟭 𝗶𝘀 𝗲𝗺𝗯𝗲𝗱𝗱𝗲𝗱 𝗯𝗲𝗹𝗼𝘄 𝗶𝗻 𝘁𝗵𝗶𝘀 𝗽𝗼𝘀𝘁. If you want the 𝗻𝗲𝘅𝘁 𝗽𝗮𝗿𝘁𝘀 (𝗣𝗮𝗿𝘁 𝟮–𝟰) and the compiled version, Share the post and DM : “𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀” #AgenticAI #AIAgents #AIEngineering #AIProducts #LLM #ProductManagement #BuildInPublic
-
Anand Arivukkarasu shared this𝗪𝗲’𝗿𝗲 𝗶𝗻 𝘁𝗵𝗲 𝗺𝗶𝗱𝗱𝗹𝗲 𝗼𝗳 𝗮 𝗳𝘂𝗻𝗱𝗮𝗺𝗲𝗻𝘁𝗮𝗹 𝘀𝗵𝗶𝗳𝘁 𝗶𝗻 𝗵𝗼𝘄 𝗦𝗮𝗮𝗦 𝗶𝘀 𝗯𝘂𝗶𝗹𝘁. For years, the formula was simple: Screens + Menus + Dashboards. The product dictated the steps. The user had to learn the system. To add value, companies just added more screens. But products like Cursor and Lovable are proving that the old model is breaking. What makes them interesting isn’t just that they use AI—it’s that they changed the interaction model entirely. Old Way: Force the user to navigate complex flows. New Way: Let the user express intent, and let agents handle the complexity. Does this mean everything becomes a chatbot? No. Pure conversation has limits. Traditional UI provides context that chat can't. The winning formula I’m seeing emerge is a hybrid approach: Conversation to capture intent. Agents to do the heavy lifting in the back Lightweight Menu based navigational UI to provide clarity and control. 𝗪𝗲 𝗮𝗿𝗲 𝗺𝗼𝘃𝗶𝗻𝗴 𝗮𝘄𝗮𝘆 𝗳𝗿𝗼𝗺 𝘁𝗲𝗮𝗰𝗵𝗶𝗻𝗴 𝘂𝘀𝗲𝗿𝘀 𝗵𝗼𝘄 𝘁𝗼 𝘂𝘀𝗲 𝘀𝗼𝗳𝘁𝘄𝗮𝗿𝗲, 𝗮𝗻𝗱 𝘁𝗼𝘄𝗮𝗿𝗱 𝘀𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝘁𝗵𝗮𝘁 𝗮𝗱𝗮𝗽𝘁𝘀 𝘁𝗼 𝗵𝗼𝘄 𝗽𝗲𝗼𝗽𝗹𝗲 𝘁𝗵𝗶𝗻𝗸. That’s the shift worth paying attention to.
-
Anand Arivukkarasu shared this𝗪𝗵𝘆 𝗔𝗜 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝘀 𝗙𝗮𝗶𝗹 𝗮𝘁 𝘁𝗵𝗲 𝗦𝗲𝗰𝗼𝗻𝗱 𝗠𝗶𝗹𝗲𝘀𝘁𝗼𝗻𝗲 — 𝗡𝗼𝘁 𝘁𝗵𝗲 𝗙𝗶𝗿𝘀𝘁 Most AI products fail not at launch. They fail at Milestone 2 — the moment after the initial “wow” disappears and users start behaving like humans again, not tourists. Here’s the core pattern I see across AI startups : 𝗠𝗶𝗹𝗲𝘀𝘁𝗼𝗻𝗲 𝟭: “𝗧𝗵𝗶𝘀 𝗶𝘀 𝗰𝗼𝗼𝗹.” Every AI agent, wrapper, or assistant clears this hurdle. Users try it. They get a result. They feel the dopamine rush of “AI magic.” Founders celebrate. Retention graphs look promising for 14–30 days. This is not success. This is novelty. Novelty behaves like a sugar high. It spikes. It always crashes. 𝗠𝗶𝗹𝗲𝘀𝘁𝗼𝗻𝗲 𝟮: “𝗪𝗶𝗹𝗹 𝗜 𝘂𝘀𝗲 𝘁𝗵𝗶𝘀 𝗲𝘃𝗲𝗿𝘆 𝘄𝗲𝗲𝗸?” This is the graveyard. This is where: “AI writing tools” go to die “AI research agents” flatten “AI copilots” get uninstalled “AI dashboards” become ghost towns “AI wrappers” vanish as quickly as they appear 𝗪𝗵𝘆? Because Milestone 2 is no longer about capability. It’s about psychology. It’s driven by: Habits Trust Emotion Reliability expectations Identity (“Does using this make me feel competent?”) Loss aversion (“Will this break if I rely on it?”) Most AI products don’t understand the human they’re serving. They understand the model they’re using. 𝗦𝗼𝗺𝗲 𝗲𝘅𝗮𝗺𝗽𝗹𝗲𝘀 : 1) The “Smart Assistant” that becomes a stranger Milestone 1: It answers your prompt well. Milestone 2: You realize it doesn’t remember context, tone, preferences, or your working style. It feels like starting over every time. AI becomes a chore. Outcome: Churn. 2) The “AI Inbox / Email Agent plugin” that overwhelms Milestone 1: It drafts replies quickly. Milestone 2: You notice: Wrong assumptions Overconfident rewrites Tone mismatches Hallucinated commitments Replies you’d never send Users don’t want speed in communication. They want safety. Outcome: Turned off within 10–14 days. 𝗛𝗲𝗿𝗲’𝘀 𝘁𝗵𝗲 𝗱𝗲𝗲𝗽𝗲𝗿 𝘁𝗿𝘂𝘁𝗵: AI products fail when they cross from novelty → dependency. Dependency requires: Predictability Emotion Behavioral reinforcement Onboarding into a new identity A sense of progress Trust compounding A stable workflow A feeling of “I can’t go back” These are psychological milestones, not technical ones. This is why product architecture matters more in AI than in traditional SaaS. It’s no longer about what the tool can do — it’s about what the user is willing to trust, adopt, and depend on. — 𝗠𝘆 𝟮 𝗰𝗲𝗻𝘁𝘀: If you want your AI products to survive Milestone 2, Go for behavior, not outputs. Solve for needs, not features Design for identity, not actions. Architect for trust, not demos. Most AI products can clear the “magic” bar. Almost none can clear the human bar. That’s where the real work starts for the AI PM!
-
Anand Arivukkarasu shared thisThanks for the good response to my recent post on the transition from Product Managers to Product Architects! As promised, here's the AI Product Architect Playbook covering key areas: ✅ AI-assisted design ideation ✅ Rapid coding and prototyping ✅ Data-driven decision making ✅ AI-powered user research ✅ Intelligent prioritization ✅ Understanding AI product architecture Feel free to share it widely to help more AI developers and Product Managers and Entrepreneurs! I'd also love to hear about other AI tools you're using in Product Development and Management. #AI #AIProductManagement #ProductManagement #ProductArchitect Mixpanel Tability ChatGPT Notion Figma Cursor Replit GitHub Microsoft Microsoft Copilot Lovable Galileo🔭 PostHog Claude AI Vizard.ai Statsig
-
Anand Arivukkarasu shared thisFrom Product Managers to Product Architects I remember 18 years ago, as a software engineer, my manager handed me a unique challenge for a product prototype: become a one-person team— design, develop, data analyze, gather & prioritize requirements—all rolled into one, the role he called Product Architect. Back then, there were no product managers and designers were rare. Fast forward to today, this multi-dimensional role is becoming the norm, empowered by AI. With AI, Product Managers are evolving into Product Architects, capable of: ✅ Ideating designs without dedicated designers ✅ Rapid coding and prototyping with AI assistance ✅ Making data-driven decisions without large data teams ✅ Conducting scaled user research using AI insights ✅ Intelligent prioritization for smarter roadmaps ✅ Understanding AI product architecture to build AI first products. I've built a simple playbook, the "AI Product Architect Playbook," detailing how PMs can seamlessly transition to this new role, leveraging various AI tools to enhance their impact. Enabling startups in the AI space, I'm exploring this transformation daily, using the latest AI tools to empower lean product teams and entrepreneurs. If you are interested in the early version of this playbook, comment "Playbook" or "AI Product Architect Playbook," and I'll share the booklet with you to get your feedback. I would like to learn about more tools that PMs are using these areas other than the ones listed below. #AI #AIProductManagement #ProductManagement #ProductArchitect Mixpanel Productboard Tability ChatGPT Notion Figma Cursor Replit GitHub Microsoft Copilot Lovable Galileo🔭 PostHog Claude AI Vizard.ai Statsig
-
Anand Arivukkarasu shared thisEvals: The Essential Skill for AI Product Development & Product Management In the fast-evolving AI landscape, writing impactful evals is critical to product success. Here's a quick guide: 🔍 What are Evals? Structured assessments measuring AI product performance and quality. Combine metrics, qualitative insights, and user feedback. 🌟 Why Evals Matter: Identify AI model strengths & weaknesses. Guide iterative improvements. Validate assumptions & hypotheses. Build user trust & enhance satisfaction. 🎯 Popular Eval Approaches: Benchmarking: Compare your AI to industry standards or competitors. User Studies: Capture qualitative user feedback for insights. A/B Testing: Experiment with variations to find the best-performing option. Automated Metrics: Use objective metrics (accuracy, precision, recall) for evaluation. 📌 Structure Your Eval Clearly: Objective: Define precisely what you're measuring. Methodology: Outline data collection & analysis methods. Results: Present concise, data-driven findings. Recommendations: Offer clear, actionable insights. 🔑 Effective Eval Workflow: Define Scope → Select Approach → Gather Data → Analyze → Draft Eval → Review & Iterate. What's your favorite eval strategy? #ProductManagement #AI #Evaluations #TechLeadership #evals #AI #ProductManagement
-
Anand Arivukkarasu reposted thisAnand Arivukkarasu reposted thisSF on Saturday morning: a large % of Waymos are ferrying solo kids (8-14 yo), presumably to sports or other activities. A friend told me that 85% of parents at their kids’ SF school use Waymo for kids pickup / drop off: earlier it was ~10% using Uber and Lyft. Shows how Waymo has grown by expanding the market, serving segments who otherwise were non-consumers of ride sharing. Though it’s expensive to build companies serving non-consumption markets, some of the best and greatest companies (SpaceX, AirBnB, etc) have and will be built creating new habits and markets.
-
Anand Arivukkarasu liked thisAnand Arivukkarasu liked thisThe wait was short! I’m excited to share that I'm joining Lovable to lead Influencer Marketing 🎉 For those who have been watching Lovable’s trajectory, you'll know that they’ve built something that puts real creative power in the hands of people who have ideas, but not necessarily the technical background to execute them. As someone who spent the last few years working with creators that are also builders, I am so excited to continue to enable creators to tell their best stories, launch something new or pursue a lifelong dream. I'm still passionate about building an influencer program that’s rooted in genuine community, not just content. If you’re a creator who’s been building with Lovable (or curious about it), want to bring something to the world, or just want to connect - please reach out! Thank you to Cecilia Stallsmith Anton Osika Sam Vinden Senka Hadzimuratovic Vikas Bhagat for bringing me into the crew! More to come ❤️
-
Anand Arivukkarasu liked thisEverything costs too much and wages haven't kept up in New Zealand. That's not bad luck. That's what happens when four banks control mortgages, two supermarkets control food, and four companies control power in a country that generates 85% of its electricity from renewables. New Zealand doesn't have a cost of living crisis. It has a competition problem. And underneath that, a productivity problem. Even the Reserve Bank's chief economist said it last week: monetary policy can't make New Zealand affordable on its own. Productivity growth is what drives real wages, and ours has lagged for decades. $1.5 trillion in residential property. $120 billion in KiwiSaver funding growth companies in New York and London while Kiwi founders can't raise capital at home. Almost nothing going into the businesses that actually create wealth. A nurse, a builder, a software engineer earns 40% more in Sydney. People aren't leaving for an OE. They're leaving because the numbers make sense. But New Zealand has real assets. Clean energy. Strong rule of law. World-class agriculture. A Māori economy worth $90 billion. A million Kiwis offshore who'd invest back home given a reason. A startup ecosystem proving Kiwi companies can win globally. The problem has never been what New Zealand has. It's that policy settings point all of it in the wrong direction: into property, into oligopoly rents, into protected markets instead of productivity and innovation. Denmark did it. The Netherlands did it. New Zealand has every ingredient they had and some they didn't. The window is now. Growth NZAnand Arivukkarasu liked thisThanks Garth Bray and NZ Herald Now for inviting me and Anna Kominik in to talk about some of the structural challenges holding NZ back in critical sectors like energy and banking, and why Growth NZ was formed to call for progress and a concrete plan to grow the economy so that all New Zealanders can thrive: https://lnkd.in/eVBeMifm If you’d like to get involved, get in touch at growthnz.org.nz/contactTech investor Maya Pan – Growth NZ – takes a swing at Australian-owned banks | Herald NOW businessTech investor Maya Pan – Growth NZ – takes a swing at Australian-owned banks | Herald NOW business
-
Anand Arivukkarasu liked thisAnand Arivukkarasu liked thisBeste leerkrachten, directie en schoolteams, Wil je jouw lessen krachtiger, duidelijker én leuker maken met eenvoudige tekeningen? Op 23 april geef ik in Kortrijk de training “Maak leren zichtbaar met sketchnoting” — speciaal ontwikkeld voor het onderwijs, samen met mensen uit het werkveld. Het is: *Praktisch *Toegankelijk (je hoeft niet te kunnen tekenen) *Meteen toepasbaar in de klas Er zijn nog een aantal plaatsen vrij — dus als dit al even op je lijstje stond: dit is je kans 👍 Inschrijven kan je hier: https://lnkd.in/eEcSm_wA #sketchnoting #dualcoding #visueeltrainen #visualstorytelling Vik Pauwels Lara Geerardyn
-
Anand Arivukkarasu liked thisAnand Arivukkarasu liked thisIs Menlo Ventures building the most impressive AI portfolio in all of venture right now? I think you could make that argument ⤵️ 2026 is already looking like a breakout year for the storied firm. Many VCs are playing the volume game - Menlo is doing the opposite: ➡️ Low volume ➡️ High conviction ➡️ Early bets And it’s compounding. Look at their AI investments: 🔸 Anthropic 🔸 Suno 🔸 Lovable 🔸 Higgsfield 🔸 Mercor 🔸 Chai Discovery 🔸 Delphi 🔸 Wispr Flow 🔸 Hale 🔸 Abnormal + many more exceptional companies (Legora, Inception, Goodfire, OpenRouter)! The strategy is simple (and rare): Only ~2 investments per partner per year. Go early. Many deals at Seed. While others “spray and pray” Menlo picks few and goes deep. And when it works… it really works. Thanks to Harmonic for the data 🫶
-
Anand Arivukkarasu liked this:)Anand Arivukkarasu liked thisIs Menlo Ventures building the most impressive AI portfolio in all of venture right now? I think you could make that argument ⤵️ 2026 is already looking like a breakout year for the storied firm. Many VCs are playing the volume game - Menlo is doing the opposite: ➡️ Low volume ➡️ High conviction ➡️ Early bets And it’s compounding. Look at their AI investments: 🔸 Anthropic 🔸 Suno 🔸 Lovable 🔸 Higgsfield 🔸 Mercor 🔸 Chai Discovery 🔸 Delphi 🔸 Wispr Flow 🔸 Hale 🔸 Abnormal + many more exceptional companies (Legora, Inception, Goodfire, OpenRouter)! The strategy is simple (and rare): Only ~2 investments per partner per year. Go early. Many deals at Seed. While others “spray and pray” Menlo picks few and goes deep. And when it works… it really works. Thanks to Harmonic for the data 🫶
-
Anand Arivukkarasu liked thisAnand Arivukkarasu liked thisWhat an incredible start to GTC 2026! Day 1 was a massive validation of the work the CoreWeave team has been doing to build the Essential Cloud for AI. From major hardware leaps to a defining shout-out from Jensen, Day 1 was one for the books. 🚀 Launched: NVIDIA HGX B300 on CoreWeave We are officially bringing the NVIDIA HGX B300 to CoreWeave Cloud. This is a game-changer for the next phase of agentic AI, offering the reasoning performance and memory density required for multi-trillion parameter models. Check out the deep dive on what this means for scaling agents in Harsh Banwait’s latest blog post 👉 https://lnkd.in/eCjqnBvt 🛠️ Real Infrastructure for Real AI Did I hear you say "real"? Chen Goldberg and Corey Sanders took the stage to show what production-scale training and inference actually look like. 1️⃣ Live Demos: We showcased the CoreWeave Mission Control Agent and the operational ease of SUNK (Slurm on Kubernetes) for provisioning. 2️⃣ We talked about CoreWeave ARENA—our production-scale AI lab that allows teams to test real workloads on purpose-built infra before they move to production. 3️⃣ The Customer Voice: A huge thanks to Sebastian Yunge from Mercado Libre for joining us to discuss how they are tackling the complexities of production AI. 🎤 The "Mic-Drop" Keynote Moment It was a proud moment for the entire team to see Jensen Huang talk about the ecosystem building our AI future. Seeing CoreWeave's AI stack showcased was a great moment. 🤝 It takes a village. Jensen emphasized the ecosystem in his talk and it truly takes a village. The show floor is teeming with amazing innovations. Come see us at booth #913 Thanks to the amazing team that makes it all happen #GTC2026 #CoreWeave #EssentialCloud
-
Anand Arivukkarasu liked thisAnand Arivukkarasu liked thisTHIS IS BIG🔥Interested in creating the semicon future at Tampere, Finland, Europe ? Grab on the FutureChips fellowship. 36 months. 20 (twenty) full time postdoc positions. Freedom to select your own research area. At least 2 months in industry. Finland ranks as "world's happiest country" for 9th year in a row. Got interested ? Prepare to apply. More information available from futurechips@tuni.fi This is not the snack version - this is a real deal. Tampere University is in the crossroads of space, WBG chips, AI, microelectronics, photonics, quantum, signal processing, medical and social sciences creating the sustainable and more secure planetary future. Stronger together including international state of art collaboration (TBA): Tampere University Tampere University of Applied Sciences Business Tampere Tampereen kaupunki - City of Tampere. Feel free to repost.
Experience & Education
-
Ideas2IT Technologies
View Anand’s full experience
See their title, tenure and more.
or
Volunteer Experience
Patents
-
Applying insights from mobile consumer data for prediction modeling
US
Courses
-
Design and Innovation
-
-
Entrepreneurship
-
-
Finance (1 and 2)
-
-
IP law and patents
-
-
International business
-
-
Investments
-
-
Management Strategy
-
-
Managerial accounting
-
-
Market research
-
-
Media management
-
-
Operations Mgmt
-
-
Pricing, Product marketing
-
-
Product Management
-
-
Risk and Crisis management
-
-
Strategy
-
-
Web intelligence and Big Data
Coursera
Honors & Awards
-
CIO Team Excellence Award
Verizon CIO
-
Above and Beyond Excellence Award
Verizon
-
Above and Beyond Excellence Award
Verizon
-
Patent award
Sprint
Patent on applying insights from mobile data to consumer products
-
i-Recognize award
Sprint
Test Scores
-
GMAT
Score: 740
Organizations
-
Adtech Meetup
Organisers
- Present -
MMA
location advertising Working group
Recommendations received
5 people have recommended Anand
Join now to viewOther similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content