Engineering

AI raised the floor. Engineering excellence raises the ceiling. It's so riveting to see new LLM models get published and the step changes that are happening. AI has made it dramatically easier to produce code. It has simultaneously made it much harder to hide weak engineering fundamentals. AI is raising the floor, meaning more people can generate software and prototypes quickly. But engineering excellence raises the ceiling: determining whether that code becomes a reliable, scalable system that actually creates enterprise value. AI is exposing something many organizations have quietly carried for years: technical debt, fragile architectures, and disconnected data foundations. When systems aren't built well, AI doesn't fix that. It simply reveals it faster. 💡  𝗜 𝗮𝗺 𝗮 𝘀𝘁𝗿𝗼𝗻𝗴 𝗯𝗲𝗹𝗶𝗲𝘃𝗲𝗿 𝘁𝗵𝗮𝘁 𝘁𝗼 𝗺𝗮𝘅𝗶𝗺𝗶𝘇𝗲 𝗔𝗜 𝘃𝗮𝗹𝘂𝗲, 𝘄𝗲 𝗻𝗲𝗲𝗱 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 𝗲𝘅𝗰𝗲𝗹𝗹𝗲𝗻𝗰𝗲. So what does engineering excellence look like right now? I think about it as four pillars: ▸ 𝗔𝗜-𝗥𝗲𝗮𝗱𝘆 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲: AI doesn't work well on top of poor architecture. Modernizing legacy code without addressing underlying structure just produces the wrong architecture faster. ▸ 𝗛𝗶𝗴𝗵-𝗤𝘂𝗮𝗹𝗶𝘁𝘆 𝗗𝗮𝘁𝗮 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻𝘀: AI is only as intelligent as the data it reasons over. You can't shortcut this layer and even a strong foundation must continuously evolve. ▸ 𝗦𝗲𝗰𝘂𝗿𝗲 𝗮𝗻𝗱 𝗢𝗯𝘀𝗲𝗿𝘃𝗮𝗯𝗹𝗲 𝗦𝘆𝘀𝘁𝗲𝗺𝘀: As AI agents become more autonomous, seeing what's happening and why becomes non-negotiable. Governance isn't just policy it's instrumentation and operationalization, as many of you noted in my last post. ▸ 𝗗𝗶𝘀𝗰𝗶𝗽𝗹𝗶𝗻𝗲𝗱 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 𝗣𝗿𝗮𝗰𝘁𝗶𝗰𝗲𝘀: Spec discipline, test rigor, strong code review, clear ownership are not legacy practices to abandon, but more important than ever. AI rewards good fundamentals and makes the consequences of weak ones more visible, faster. There's a real shift in how engineers spend their time. Less writing foundational code. More orchestrating systems: designing architecture, shaping how AI agents interact, validating outputs with genuine judgment. I see our senior engineers flying because their systems thinking depth makes AI a true force multiplier. Earlier-career engineers are learning, but need more deliberate mentorship than ever. When AI can simulate senior output, the risk is gaining confidence without gaining understanding. The best thing leaders can do: create conditions where engineers are proud of how they build, not just what they ship. The time savings alone aren't the win. For us, we are investing in deeper architecture work, stronger data foundations, the next generation of agentic capabilities and I believe that's the winning combo. 𝗗𝗼 𝘆𝗼𝘂 𝗮𝗴𝗿𝗲𝗲 𝘁𝗵𝗮𝘁 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 𝗲𝘅𝗰𝗲𝗹𝗹𝗲𝗻𝗰𝗲 𝗶𝘀 𝗺𝗼𝗿𝗲 𝗶𝗺𝗽𝗼𝗿𝘁𝗮𝗻𝘁 𝘁𝗵𝗮𝗻 𝗲𝘃𝗲𝗿?

The world's largest sand battery has been inaugurated in Finland. Developed by Polar Night Energy, this high-temperature thermal energy storage system stores heat in sand using low-cost, clean electricity. The project is a powerful example of how thermal storage can enhance grid flexibility, decarbonise heating, and accelerate the energy transition. - It can store up to 100 MWh of thermal energy. - It has a round-trip efficiency of 90%. - It offers a cost-effective alternative to lithium-ion batteries for long-term heat storage. - By replacing an old woodchip plant, the sand battery is expected to cut the local heating network's carbon emissions by 70%.

Roadmap to Learn Agentic AI This roadmap breaks down the journey into 12 focused stages: – Grasp the core differences between traditional AI and autonomous agents – Build a solid foundation in ML, LLMs, and frameworks like LangGraph, CrewAI, and AutoGen – Understand how agents use memory, plan actions, and collaborate – Learn to implement retrieval-augmented generation (RAG) and adaptive reinforcement learning – Deploy agents in real-world scenarios with performance monitoring and continuous improvement If you're building AI that goes beyond chat interfaces, this roadmap will help you architect systems that are capable, contextual, and action-oriented. Feel free to save or share if you find it valuable.

We’re planting trees — but losing biodiversity. Global efforts to restore forests are gathering pace, driven by promises of combating climate change, conserving biodiversity, and improving livelihoods. Yet a recent paper published in Nature Reviews Biodiversity warns that the biodiversity gains from these initiatives are often overstated — and sometimes absent altogether. Forest restoration is at the heart of Target 2 of the Kunming-Montreal Global Biodiversity Framework, which aims to place 30% of degraded ecosystems under effective restoration by 2030. But the gap between ambition and outcome is wide. "Biodiversity will remain a vague buzzword rather than an actual outcome" unless projects explicitly prioritize it, the authors caution. Restoration has typically prioritized utilitarian goals such as timber production, carbon sequestration, or erosion control. This bias is reflected in the widespread use of monoculture plantations or low-diversity agroforests. Nearly half of the Bonn Challenge’s forest commitments consist of commercial plantations of exotic species — a trend that risks undermining biodiversity rather than enhancing it. Scientific evidence shows that restoring biodiversity requires more than planting trees. Methods like natural regeneration — allowing forests to recover on their own — can often yield superior biodiversity outcomes, though they face social and economic barriers. By contrast, planting a few fast-growing species may sequester carbon quickly but offers little for threatened plants and animals. Biodiversity recovery is influenced by many factors: the intensity of prior land use, the surrounding landscape, and the species chosen for restoration. Recovery is slow — often measured in decades — and tends to lag for rare and specialist species. Alarmingly, most projects stop monitoring after just a few years, long before ecosystems stabilize. However, the authors say there are reasons for optimism. Biodiversity markets, including emerging biodiversity credit schemes and carbon credits with biodiversity safeguards, could mobilize new financing. Meanwhile, technologies like environmental DNA sampling, bioacoustics, and remote sensing promise to improve monitoring at scale. To turn good intentions into reality, the paper argues, projects must define explicit biodiversity goals, select suitable methods, and commit to long-term monitoring. Social equity must also be central. "Improving biodiversity outcomes of forest restoration… could contribute to mitigating power asymmetries and inequalities," the authors write, citing examples from Madagascar and Brazil. If designed well, forest restoration could help address the twin crises of biodiversity loss and climate change. But without a deliberate shift, billions of dollars risk being spent on projects that plant trees — and little else. 🔬 Brancalion et al (2025): https://lnkd.in/gG6X36WP

Scaling from 50 to 100 employees almost killed our company. Until we discovered a simple org structure that unlocked $100M+ in annual revenue. In my 10+ years of experience as a founder, one of the biggest challenges I faced in scaling was bridging the organizational gap between startup and enterprise. We hit that wall at around 100~ employees. What worked beautifully with a small team suddenly became our biggest obstacle to growth. The problem was our functional org structure: Engineers reporting to engineering, product to product, business to business. This created a complex dependency web: • Planning took weeks • No clear ownership  • Business threw Jira tickets over the fence and prayed for them to get completed • Engineers didn’t understand priorities and worked on problems that didn’t align with customer needs That was when I studied Amazon's Single-Threaded Owner (STO) model, in which dedicated GMs run independent business units with their own cross-functional teams and manage P&L It looked great for Amazon's scale but felt impossible for growing companies like ours. These 2 critical barriers made it impractical for our scale: 1. Engineering Squad Requirements: True STO demands complete engineering teams (including managers) reporting to a single owner. At our size, we couldn't justify full engineering squads for each business unit. To make it work, we would have to quadruple our engineering headcount. 2. P&L Owner Complexity: STO leaders need unicorn-level skills: deep business acumen and P&L management experience. Not only are these leaders rare and expensive, but requiring all these skills in one person would have limited our talent pool and slowed our ability to launch new initiatives. What we needed was a model that captured STO's focus and accountability but worked for our size and growth needs. That's when we created Mission-Aligned Teams (MATs), a hybrid model that changed our execution (for good) Key principles: • Each team owns a specific mission (e.g., improving customer service, optimizing payment flow) • Teams are cross-functional and self-sufficient,  • Leaders can be anyone (engineer, PM, marketer) who's good at execution • People still report functionally for career development • Leaders focus on execution, not people management The results exceeded our highest expectations: New MAT leads launched new products, each generating $5-10M in revenue within a year with under 10 person teams. Planning became streamlined. Ownership became clear. But it's NOT for everyone (like STO wasn’t for us) If you're under 50 people, the overhead probably isn't worth it. If you're Amazon-scale, pure STO might be better. MAT works best in the messy middle: when you're too big for everyone to be in one room but too small for a full enterprise structure. image courtesy of Manu Cornet ------ If you liked this, follow me Henry Shi as I share insights from my journey of building and scaling a  $1B/year business.

Exciting updates on Project GR00T! We discover a systematic way to scale up robot data, tackling the most painful pain point in robotics. The idea is simple: human collects demonstration on a real robot, and we multiply that data 1000x or more in simulation. Let’s break it down: 1. We use Apple Vision Pro (yes!!) to give the human operator first person control of the humanoid. Vision Pro parses human hand pose and retargets the motion to the robot hand, all in real time. From the human’s point of view, they are immersed in another body like the Avatar. Teleoperation is slow and time-consuming, but we can afford to collect a small amount of data.  2. We use RoboCasa, a generative simulation framework, to multiply the demonstration data by varying the visual appearance and layout of the environment. In Jensen’s keynote video below, the humanoid is now placing the cup in hundreds of kitchens with a huge diversity of textures, furniture, and object placement. We only have 1 physical kitchen at the GEAR Lab in NVIDIA HQ, but we can conjure up infinite ones in simulation. 3. Finally, we apply MimicGen, a technique to multiply the above data even more by varying the *motion* of the robot. MimicGen generates vast number of new action trajectories based on the original human data, and filters out failed ones (e.g. those that drop the cup) to form a much larger dataset. To sum up, given 1 human trajectory with Vision Pro  -> RoboCasa produces N (varying visuals)  -> MimicGen further augments to NxM (varying motions). This is the way to trade compute for expensive human data by GPU-accelerated simulation. A while ago, I mentioned that teleoperation is fundamentally not scalable, because we are always limited by 24 hrs/robot/day in the world of atoms. Our new GR00T synthetic data pipeline breaks this barrier in the world of bits. Scaling has been so much fun for LLMs, and it's finally our turn to have fun in robotics! We are creating tools to enable everyone in the ecosystem to scale up with us: - RoboCasa: our generative simulation framework (Yuke Zhu). It's fully open-source! Here you go: http://robocasa.ai - MimicGen: our generative action framework (Ajay Mandlekar). The code is open-source for robot arms, but we will have another version for humanoid and 5-finger hands: https://lnkd.in/gsRArQXy - We are building a state-of-the-art Apple Vision Pro -> humanoid robot "Avatar" stack. Xiaolong Wang group’s open-source libraries laid the foundation: https://lnkd.in/gUYye7yt - Watch Jensen's keynote yesterday. He cannot hide his excitement about Project GR00T and robot foundation models! https://lnkd.in/g3hZteCG Finally, GEAR lab is hiring! We want the best roboticists in the world to join us on this moon-landing mission to solve physical AGI: https://lnkd.in/gTancpNK

Should you try Google’s famous “20% time” experiment to encourage innovation? We tried this at Duolingo years ago. It didn’t work. It wasn’t enough time for people to start meaningful projects, and very few people took advantage of it because the framework was pretty vague. I knew there had to be other ways to drive innovation at the company. So, here are 3 other initiatives we’ve tried, what we’ve learned from each, and what we're going to try next. 💡 Innovation Awards: Annual recognition for those who move the needle with boundary-pushing projects. The upside: These awards make our commitment to innovation clear, and offer a well-deserved incentive to those who have done remarkable work. The downside: It’s given to individuals, but we want to incentivize team work. What’s more, it’s not necessarily a framework for coming up with the next big thing. 💻 Hackathon: This is a good framework, and lots of companies do it. Everyone (not just engineers) can take two days to collaborate on and present anything that excites them, as long as it advances our mission or addresses a key business need. The upside: Some of our biggest features grew out of hackathon projects, from the Duolingo English Test (born at our first hackathon in 2013) to our avatar builder. The downside: Other than the time/resource constraint, projects rarely align with our current priorities. The ones that take off hit the elusive combo of right time + a problem that no other team could tackle. 💥 Special Projects: Knowing that ideal equation, we started a new program for fostering innovation, playfully dubbed DARPA (Duolingo Advanced Research Project Agency). The idea: anyone can pitch an idea at any time. If they get consensus on it and if it’s not in the purview of another team, a cross-functional group is formed to bring the project to fruition. The most creative work tends to happen when a problem is not in the clear purview of a particular team; this program creates a path for bringing these kinds of interdisciplinary ideas to life. Our Duo and Lily mascot suits (featured often on our social accounts) came from this, as did our Duo plushie and the merch store. (And if this photo doesn't show why we needed to innovate for new suits, I don't know what will!) The biggest challenge: figuring out how to transition ownership of a successful project after the strike team’s work is done. 👀 What’s next? We’re working on a program that proactively identifies big picture, unassigned problems that we haven’t figured out yet and then incentivizes people to create proposals for solving them. How that will work is still to be determined, but we know there is a lot of fertile ground for it to take root. How does your company create an environment of creativity that encourages true innovation? I'm interested to hear what's worked for you, so please feel free to share in the comments! #duolingo #innovation #hackathon #creativity #bigideas

👁 Imagine losing your sight for 10 years… and then, the very first thing you do is recognize the faces of your loved ones again. That’s what happened to Jamal Furani, 78, thanks to a breakthrough in medical innovation: a fully synthetic cornea implant. No donor tissue. No immune rejection. A device that integrates directly with the eye’s own tissue. 💡 The deeper insight: The true revolution here isn’t only technological. It’s structural. Today, corneal blindness affects millions worldwide, but most can’t be treated because there simply aren’t enough donor corneas. A synthetic cornea changes the equation. It turns a scarce resource (donations) into a potentially unlimited one (innovation). And here’s what few realize: this implant doesn’t just restore vision. It restores autonomy, dignity, and human connection. Those are the “side effects” that make technology truly transformative. 👉 My take: The future of medicine won’t just be about “healing.” It will be about reinventing our organs — sometimes with solutions even better than the originals. If you could enhance or replace one organ with technology, which would you choose first? #Healthcare #Innovation #Biotech #FutureOfMedicine

This is how Anthropic decides what to build next—and it's brilliant. Instead of endless spec documents and roadmap debates, the Claude Code team has cracked the code on feature prioritization: prototype first, decide later. Here's their process (shared by Catherine Wu, Product Lead at Anthropic): Step 1: Idea → Prototype Got a feature idea? Skip the spec. Build a working prototype using Claude Code instead. Step 2: Internal Launch Ship that prototype to all Anthropic engineers immediately. No polish required—just functionality. Step 3: Watch & Listen Track usage religiously. Collect feedback actively. Let real behavior, not opinions, guide decisions. Step 4: Data-Driven Prioritization - High usage + positive feedback → roadmap priority - Low engagement or complaints → back to iteration This "prototype-first product shaping" flips traditional product development on its head. Instead of guessing what users want, they're measuring what users actually use. The beauty? They're dogfooding their own tool to build their own tool. The feedback loop is immediate, honest, and impossible to ignore. The takeaway: Your best product decisions come from real user behavior, not theoretical frameworks. Sometimes the fastest way to validate an idea isn't a survey or interview—it's a working prototype.