Ambrish Verma

We chose a good subject for our march Autotech Council meeting, and after this year's CES, it seems even more obvious. Our meeting on the AI Defined Vehicle will show how it's a natural next step after the Software Defined Vehicle phase that we've been in Tesla launched the Model S in 2012. Now, most carmakers have SDVs. It's time to layer on the AI in user-facing roles, and in invisible functions managing the vehicle.

1

💡 Startups are finding unexpected, creative ways to apply AI across industries you might not imagine. We’ll spotlight 25 innovative use cases that reveal the breadth and depth of what’s possible when bold founders meet cutting-edge technology. 🔗 https://bit.ly/46ttjmM

41

2 Comments

Most of the AI tooling conversation right now centers on code generation. Cognition, Cursor, Claude Code....these tools are genuinely useful, but engineers are drowning in production issues. At our latest Greylock Partners Change Agents event, Mayank Agarwal and Rushin Shah from Resolve AI walked through a reality most engineering teams live through: with AI coding tools "you can crank out code much faster, but teams aren't really shipping to production that much faster." Some takeaways from our conversation: 1/ The real problem isn't code generation. Mayank saw this at Splunk where engineers spent only 10-20% of their time writing new features. The rest was fighting incidents, navigating tribal knowledge across hundreds of microservices, and operating tools that don't talk to each other. 2/ Production systems are different from code. Your codebase is static. Production is emergent behavior from infrastructure, deployments, traffic patterns, to tribal knowledge buried in Slack. As Mayank stated, the issue is "a coding agent cannot predict these things without the deep understanding of what's actually happening in production." 3/ The scale required is massive. Resolve is now consuming a trillion tokens per quarter—1% of all Azure usage. That's running Stack Overflow from scratch twice per week. "Vibe debugging" is the missing piece, Rushin's term for getting clarity on production systems the way vibe coding accelerates writing code. Check out the full episode in the comments

104

28 Comments

🚀 Act Three of my Hypermodern GenAI series is live! This chapter dives into LangChain’s Deep Agents and what they mean for the future of agent frameworks. From shallow vs deep agents to reasoning LLMs, I explore how these patterns shape the way we actually build and ship GenAI apps. The series isn’t about chasing hype — it’s about asking the hard questions: 🔹 When should we follow the defaults? 🔹 When does “depth” help, and when does it just add overhead? 🔹 How do we balance user experience, developer productivity, and system cost? Think of it as a guidebook for hypermodern generative AI agents: rethinking the playbook, experimenting carefully, and building with the user in mind.

3

1 Comment

Honored to have been invited to fal and Henry Shi’s Lean AI Leaderboard Celebration Dinner, celebrating some of the leanest and fastest-growing startups in SF. Here are some of the most overwhelming trends that were discussed among high-growth companies working with small, high-output teams: 1/ Headcount is no longer used as a success metric. As Henry Ward said, “Hiring means we failed to execute and need help.” Headcount is not a flex measure anymore. People aren’t asking “How big is your team?” anymore. They’re asking “How many customers do you have?” 2/ They're finding ways to leverage AI across every function. Every engineer uses Cursor or Windsurf. Every GTM leader is in Clay. The most efficient teams are built AI-first across the board. Thank you to Henry Shi, Burkay Gur, and Gorkem Yurtseven for putting this all together!

67

If you're in the CPG space, you're probably feeling the pressure to innovate quickly—without sacrificing quality or margin. My colleague Simran Suri dives into what an AI-powered Operating System for Product Innovation could look like—and why brands need one now. From consumer spend to purchasing patterns, inventory levels to formulation R&D—she outlines how connecting these data streams in real time can transform how products are built. 👀 Worth a read if you're thinking about scaling efficiently, staying competitive, or simply curious about the future of CPG ops. #AI #CPG #Innovation #ProductDevelopment #ConsumerBrands #Startups #DTC #DataDriven https://lnkd.in/gk9av-hN

6

1 Comment

Rob Petrosino and the PeakActivity team are up to some pretty incredible things on the AI front, as Gartner noticed recently. The funny thing is, we’re moving a bit too fast to tell all the stories. Some are locked behind NDAs, others are still inside boardrooms and innovation labs, and a few haven't even been named yet. But if you're a leader sitting there trying to decide “which AI partner is actually going to help us move — not just talk?” This is for you. We’re building AI that gets CFO approval. Most agencies talk about AI like a future state. We implement it now — responsibly, securely, and tied to business outcomes. Here’s what real AI work looks like in the wild: ✅ Turning 4-hour manual review tasks into sub-30 seconds using AI validation agents ✅ Cutting onboarding time by 60% for a global B2B manufacturer with AI workflow orchestration ✅ Automating data prep that used to block analytics teams for weeks ✅ Deploying intelligent search that feels like talking to your best product expert ✅ Creating AI copilots that augment teams rather than replace them No smoke. No hype. No buzzwords. Just teams doing the work and delivering value. If you’re serious about AI — meaning you’re ready to move with purpose, avoid the hype traps, and build something real — you’re the kind of person we want to build with. We won't promise easy. We will promise momentum. If you're ready to start your AI journey with a team that’s actually shipping — let’s talk.

3

here is the next paper in the generative retrieval research in our research team. Having just come back from #recsys conference, I know many of recommender system researchers are now using Semantic IDs to integrate ranking with large models.

105