Ben Serridge

The magic of custom silicon happens at microscopic scale. Our AWS Trainium chips built for AI training and inference are architectural marvels, each capable of trillions of calculations per second with billions of transistors working in perfect harmony. Check out this interactive feature that explains how our chip designers think like city planners, optimizing every nanometer to move data at near light speed. https://lnkd.in/gCajddmg This attention to detail lets us deliver the performance and efficiency our customers need for their most demanding AI workloads. 

1,484

43 Comments

AI is everywhere. Real transformation is still rare. We just released a paper with MIT Sloan Management Review Brasil. The title: "Culture Eats AI for Breakfast." It starts with an uncomfortable number: 95% of organizations see no measurable return from their AI investments. Not because the technology failed. Because we failed to change how we work. After 6+ years at CI&T, this is what I keep seeing: companies buy tools, spin up pilots, run hackathons, then wonder why nothing scales. The answer is almost never technical. It's cultural. It's structural. It's human. At CI&T, we don't treat AI as a side initiative. We treat it as structural change. That means defining where AI truly matters, across value streams, not isolated use cases. Redesigning the engineering cycle with human+AI collaboration built in. Connecting strategy, governance, and architecture into one operating model. And orchestrating the ecosystem — so platforms, agents, and partners work as one. Because AI doesn't scale in fragments. It scales when culture, process, and leadership align. One line from Cesar Gon's preface stayed with me: "The gap between what's possible and what's wise is widening. Faster than our courage." That's an Asimov riff. And it's never been more true. Culture eats AI for breakfast. If you're leading AI transformation, or stuck in the middle of one, this is worth your time. #AITransformation #Culture #Leadership

42

3 Comments

Excited to have João (Joe) Moura, CEO of CrewAI, join us at Observe this month in San Francisco! João has been at the forefront of a new era in AI: designing structured, role-based agents that collaborate to solve complex tasks. With CrewAI, he’s helped shape a leading open-source framework for building multi-agent systems that are modular, controllable, and production-ready. His session, “Going from 0 to Production with AI Agents,” will cover things like: ✔️ How to design agentic workflows for real use cases ✔️ What it takes to go beyond demos to actual deployment ✔️ Why testing, iteration, and observability are essential at every step The focus on real-world success stands out here. That is, how can agents communicate, make decisions, and adapt while still being steerable, measurable, and safe? So if you find yourself asking questions like: ❓How can we reliably evaluate agent performance in open-ended, real-world environments? ❓Where do coordination failures or goal misalignments tend to emerge—and how can we catch them early? ❓How do we maintain oversight, safety, and control as agent systems scale in complexity? ..then you won't want to miss this conversation. Tickets 👉 arize.com/observe-2025

55

5 Comments

TL;DR - Tax code changes today are likely to lead to more hiring for R&D roles in the US, particularly for profitable high growth companies and big tech. Predictions: - Fewer layoffs in 2026 - Hiring for IC engineers recovers first - Trend towards off-shoring R&D slows, but does not stop or reverse. India and LatAm labor is still much cheaper and lots of great talent in those geos. https://lnkd.in/gyVqV6-j

20

I've worked for many startups over the years, and many times reporting directly to VP/President/CEOs. Particularly in startups with founders as the heads, this article very clearly explains how you should approach working with leadership in these types of organizations. This is terrific to read if you're currently interviewing. https://lnkd.in/g-Np-J7u

3

Prototyping is easy, production is where the real work begins. Anyone can spin up a demo, but scaling agentic systems to enterprise standards means building for reliability, security, and trust. That’s the difference between showing what’s possible and delivering lasting value. The industry doesn’t need more experiments. It needs systems that can handle production-grade reality. That’s why I’m excited about the new CrewAI × DeepLearning.AI course with Andrew Ng. it breaks down how to take agents from concept to enterprise-ready deployment. #EnterpriseAI #AgentInfrastructure #CrewAI

5

1 Comment

This series of posts has lot's of good insights on the role of product managers as creators and the impact of AI on acceleration and exercise of that muscle, worth following.... https://lnkd.in/dE29YD8p

4

At first, people think AI makes design more "robotic," but it’s doing the opposite. It’s forcing us to be more organic. It’s time to put the 'Art' back in Product Design - thanks Amanda Elyss, Henrique Sanches and Thiago Waddington Achatz for a great event. Primeiro de muitos, sucesso Anna Calmon Sobral!

17

2 Comments

Really excited about all of the amazing work we announced yesterday. These announcements represent our first step in moving Android from an operating system into a truly intelligent system.

51

3 Comments