Vasco Pedro

Every once in a while the paradigm shift we're seeing with AI comes into sharp focus for me. Today was one of those moments when Pedro, one of our brilliant engineers, demoed the internationalization features he's been building. If you've every had to deal with internationalization you know how painful it can be. We had a non-trivial consumer-facing product translated into 3 languages in a week. Pedro subsequently built an entire translation management console that lets us auto-translate to any language (seen below is a non-validated first pass in Hangul) and then included a human-in-the-loop feature to effectively crowd source refinements. We're using AI for the translation but also as a coding assistant and UX designer to deploy features like this that once wouldn't have be worth the effort.

63

5 comentários

OneMAI has been moving a alot lately, my co-founders and I have been deep in the trenches these past few months. Let me share a few hard truths I have come to learn, not from reports, but from actually trying to build something here in Portugal. Your network will open more doors than your idea ever will. In Europe and especially Portugal, connections still count more than innovation. The right introduction often moves things faster than a killer pitch deck. Bureaucracy will humble your hustle. From opening a business account to getting access to APIs or registering with social security, processes can drag on way longer than you expect, often 2–3 times longer than what it would take in the US or UK. Here are some things I wish someone told me: - 90% of founders in Portugal earn under €1,500/month in their first two years. (Startup Portugal, 2023) - Banking integration is slow; many banks still won’t give dev API access until you have been profitable or running for a year. - Despite the innovation talk, early-stage investors still tend to back familiar industries like tourism, real estate, food. I am not saying this to scare anyone. I'm saying it because it’s real, and we don’t talk about it enough. But here is the flip side: 🇵🇹 Portugal is a strategic gateway to the EU. The talent is strong, the cost of building is low, and if you can crack the system, the upside is big. What is my takeaway so far? You will need more than an idea, you will need navigation, network, and a whole lot of patience. Have you felt any of this too? Let’s trade notes or shoot me a message. #Entrepreneurship #PortugalStartups #Road2WebSummit #HardTruths

36

10 comentários

My co-founder Diogo Bragança wrote a piece for PropertyCasualty360 on what OpenClaw (the viral and impressive AI agent everyone's been talking about) means for insurance agencies. Every few months, a new AI tool goes viral and agency owners start asking: "Should we be worried about this? Should we be using it? Is this going to change how we do business?" Fair questions. The pace of change right now is fast enough to make anyone running an operation with sensitive client data and compliance obligations feel uneasy. Diogo breaks down what's real, what's hype, and what agency owners should actually be paying attention to. Link in the comments.

33

1 comentário

This weekend, ChatGPT told us we were having a girl. We had the NIPT results in Portuguese. We didn't want to know the sex. Until two hours later, we did. So we asked ChatGPT to check if it was in the document. It was confident. It said: girl. Two hours of planning and excitement later, I thought: let me double-check in Copilot. Copilot said: "I can't find that information in this document." Same file. Completely different answer. This is what I demo in every workshop: Confident ≠ correct. The tool will give you an answer even when there is no answer. It won't flag uncertainty. It won't say "I'm guessing." It delivers the wrong result with the same confidence as the right one. I see this with teams every week: - They paste a document into ChatGPT and trust the summary without reading it - They get a confident number from Copilot and put it in a board deck - They skip verification because the output "looked right" The risk isn't that AI gets things wrong. The risk is that it gets things wrong and you don't notice. P.S. We still don't know. That's probably better. We'll wait until the doctor tells us.

100

34 comentários

Portugal Is Quietly Building a Powerful Unicorn Ecosystem Over the last few years, Portugal has emerged as one of Europe’s most exciting tech hubs driven by visionary founders, strong engineering talent, and a growing culture of innovation. From SaaS to AI, FinTech to HealthTech, these unicorns are redefining what it means to build global companies from Portugal: Top Unicorn Startups in Portugal OutSystems -  Carlos Alves | Paulo Rosado Talkdesk - Tiago Paiva | Cristina Fonseca  Sword Health - Virgílio (“V”) Bento | Márcio Colunas | Fernando Correia, MD PhD | André Eiras dos Santos Remote - Job van der Voort | Marcelo Lebre Anchorage Digital - Diogo Mónica | Nathan McCauley Aptoide - Paulo Trezentos | Alvaro Pinto Sensei - Vasco Portugal | Joana Rafael | Nuno Moutinho | Paulo Carreira Barkyn - Andre Jordão | Ricardo Macedo Feedzai -  Nuno Sebastiao | Pedro Bizarro| TEKEVER - Ricardo Mendes | Pedro Sinogas | Vitor Cristina Portugal is no longer just a beautiful destination it’s a strategic launchpad for the next wave of global innovation. #PortugalStartups #Unicorns #Innovation #TechEcosystem #Startups #VentureCapital #EuropeanTech #PortugalTech #SaaS #AI #FinTech #ScaleUp

206

1 comentário

🧬 How BioLinker is Revolutionizing AI-Driven Biomanufacturing Meeting the Future at Web Summit 2025 The real magic of Web Summit often happens after the event. While the 2025 conference was incredible, my most exciting discovery actually came through the Web Summit HealthTech Startup WhatsApp group shortly after we all left Lisbon. I had the pleasure of connecting with the team behind BioLinker. They aren't just participating in the biotech revolution—they are driving it. 🧪 The Challenge: Speed and Cost Traditional synthetic biology has a major bottleneck: it is slow and expensive. Developing new proteins or microbial strains typically involves years of trial and error in "wet labs." Time: Average R&D cycles can take 5–10 years. Cost: High investment ($5m+) with no guarantee of success. Waste: The process is resource-heavy and often relies on inefficient guessing games. 🤖 The Solution: "bAIo" Technology BioLinker changes the game by fusing Artificial Intelligence with biology. They call it their bAIo platform. Instead of relying solely on physical experiments, they use a "dry lab" approach first. Their proprietary algorithms model DNA and proteins computationally before they ever touch a test tube. This creates an instant feedback loop that: Reduces development time by 50% (from years to months). Cuts costs by half while increasing efficiency. Eliminates 90% of traditional trial-and-error waste. In one case study for luciferase production, they reduced R&D time from months to just hours, increasing enzyme production by 5x. 🚀 What Makes Them Unique? BioLinker isn't just another biotech firm. They possess a unique database built on Amazonian biodiversity. This gives them access to biological data that no other competitor has. Combined with their ability to automate the entire process—from digital design to physical production—they offer a scalable solution that big pharma and agriculture industries desperately need. 👩‍🔬 The Team Behind the Magic Technology is only as good as the people building it. BioLinker is led by an impressive duo: Mona Oliveira, PhD Oliver (CEO): A dual PhD in Biochemistry and Nanotechnology, and a winner of the Google Black Founder Grant. Phelipe Vitale (CSO): A PhD in Biochemistry with over 7 years of experience engineering microbes. 🤝 Synergies with YOUR R&D™ This is where things get exciting. At YOUR R&D, we specialize in building the robust software infrastructures that power innovation. BioLinker is bridging the gap between biological data and software. Potential synergies include: Data Infrastructure: Scaling the backend required to process their massive biodiversity datasets. AI Model Optimization: Collaborating on the software architecture that runs their predictive models. Platform UX: enhancing the user interface for their clients to visualize protein models more effectively. #HealthTech #Biotech #AIDrivenBiomanufacturing #SoftwareDevelopment #Innovation #WebSummit2025 #YOURRnD

12

1 comentário

𝐅𝐫𝐨𝐦 𝐏𝐨𝐫𝐭𝐮𝐠𝐚𝐥 𝐆𝐫𝐢𝐭 🇵🇹 𝐭𝐨 𝐆𝐥𝐨𝐛𝐚𝐥 𝐀𝐦𝐛𝐢𝐭𝐢𝐨𝐧 Let's get real about bootstrapping in Portugal: the raw challenges of building from a vibrant but under-the-radar ecosystem, and the resilience it takes to scale globally. As FULLVENUE edges closer to #Clustie's scale, this story hits close to home. Bootstrapping isn't glamorous, it's emotionally tough, with lean budgets, and betting on yourself until your story sings loud enough for VCs to listen. In Portugal, we started with that classic fork in the road: run a steady, low-profit consultancy (safe, but capped) or chase a high-potential plug&play product rocket like #Clustie (risky, but world-changing). We chose the latter. Early rounds? All about the vision, painting a picture of AI that turns fragmented first-party data into precision audience clusters, slashing waste in paid ad spends. 𝐁𝐮𝐭 𝐚𝐬 𝐰𝐞 𝐡𝐢𝐭 𝐥𝐚𝐭𝐞𝐫 𝐬𝐭𝐚𝐠𝐞𝐬, 𝐭𝐫𝐚𝐜𝐭𝐢𝐨𝐧 𝐛𝐞𝐜𝐚𝐦𝐞 𝐤𝐢𝐧𝐠: 𝐩𝐫𝐨𝐯𝐞𝐧 𝐆𝐓𝐌, 𝐜𝐥𝐢𝐞𝐧𝐭 𝐰𝐢𝐧𝐬, 𝐚𝐧𝐝 𝐦𝐞𝐭𝐫𝐢𝐜𝐬 𝐭𝐡𝐚𝐭 𝐬𝐜𝐫𝐞𝐚𝐦 𝐬𝐜𝐚𝐥𝐚𝐛𝐢𝐥𝐢𝐭𝐲. Europe's fundraising scene in 2025? Tougher than ever. Yet, that's our edge: 𝐏𝐨𝐫𝐭𝐮𝐠𝐚𝐥'𝐬 𝐭𝐚𝐥𝐞𝐧𝐭 𝐩𝐨𝐨𝐥 𝐥𝐞𝐭 𝐮𝐬 𝐛𝐮𝐢𝐥𝐝 𝐥𝐞𝐚𝐧 𝐚𝐧𝐝 𝐦𝐞𝐚𝐧, 𝐩𝐫𝐨𝐯𝐢𝐧𝐠 𝐫𝐞𝐬𝐢𝐥𝐢𝐞𝐧𝐜𝐞 𝐛𝐞𝐟𝐨𝐫𝐞 𝐰𝐞 𝐬𝐜𝐚𝐥𝐞. Our north star? Cracking Spain's powerhouse economy with #Clustie. Brands here are bleeding cash on inefficient paid ads: third-party cookies are ghosts, privacy regs are ironclad, and ROI's a guessing game. We solve the real pains, optimizing ad investments with AI that clusters audiences from owned data, 𝐛𝐨𝐨𝐬𝐭𝐢𝐧𝐠 𝐫𝐞𝐭𝐮𝐫𝐧𝐬 2-3𝐱 𝐰𝐢𝐭𝐡𝐨𝐮𝐭 𝐛𝐥𝐚𝐜𝐤 𝐦𝐚𝐠𝐢𝐜 𝐨𝐫 𝐬𝐩𝐫𝐚𝐲 & 𝐩𝐫𝐚𝐲. It's not just tech; it's the multiplier eCommerce giants need to thrive in a cookieless and overcrowded world. I'm posting this from #Sevilla, soaking in the energy at Al-Andalus Innovation Venture. This hotspot for startups, scaleups and open innovation is gold. Shouts to the incredible pros here, some with our shared track record from past collabs. What's your bootstrap battle cry? Drop a 🌍 in the comments with your toughest fundraising hurdle, let's swap stories. #FULLVENUE #AgenticClustie #AI #YourBrandLLM #PocketMate #Bootstrapping #IberianVC 𝘗𝘩𝘰𝘵𝘰: 𝘊𝘩𝘢𝘵𝘵𝘪𝘯𝘨 𝘪𝘯𝘯𝘰𝘷𝘢𝘵𝘪𝘰𝘯 𝘸𝘪𝘵𝘩 𝘙𝘢𝘧𝘢𝘦𝘭 𝘢𝘵 𝘈𝘭-𝘈𝘯𝘥𝘢𝘭𝘶𝘴 𝘐𝘯𝘯𝘰𝘷𝘢𝘵𝘪𝘰𝘯 𝘝𝘦𝘯𝘵𝘶𝘳𝘦 𝘪𝘯 𝘚𝘦𝘷𝘪𝘭𝘭𝘢. 𝘉𝘪𝘨 𝘱𝘳𝘰𝘱𝘴 𝘵𝘰 𝘩𝘪𝘮 𝘧𝘰𝘳 𝘩𝘪𝘴 𝘩𝘢𝘳𝘥𝘸𝘰𝘳𝘬 𝘢𝘯𝘥 𝘳𝘦𝘭𝘦𝘯𝘵𝘭𝘦𝘴𝘴 𝘥𝘳𝘪𝘷𝘦 𝘵𝘰 𝘭𝘦𝘢𝘳𝘯 𝘦𝘷𝘦𝘳𝘺 𝘴𝘪𝘯𝘨𝘭𝘦 𝘥𝘢𝘺. 𝘔𝘰𝘮𝘦𝘯𝘵𝘴 𝘭𝘪𝘬𝘦 𝘵𝘩𝘦𝘴𝘦 𝘧𝘶𝘦𝘭 𝘵𝘩𝘦 𝘨𝘳𝘪𝘯𝘥 𝘧𝘳𝘰𝘮 𝘗𝘰𝘳𝘵𝘶𝘨𝘢𝘭 𝘵𝘰 𝘵𝘩𝘦 𝘸𝘰𝘳𝘭𝘥.

92

1 comentário

The real bottleneck for AI in regulated industries isn't code generation. Everyone's talking about AI making developers 10x faster: Generate code quicker. Ship features sooner. That's real—and valuable. But in industries where software is the product (automotive, aerospace, medical devices, defense), the bottleneck was never just writing the code. It's proving the code is safe to ship. Verification. Traceability. Audit evidence. Compliance artifacts. That's where releases actually stall — whether it's ISO 26262 in automotive, DO-178C in aerospace, IEC 62304 for medical devices, or FDA 21 CFR Part 11. Recent data backs this: Sonar's 2026 State of Code Developer Survey reveals a massive "verification gap"—96% of developers don't fully trust AI-generated code is functionally correct, yet only 48% always verify it before committing, and 38% say reviewing AI code takes more effort than human-written code. In regulated spaces, AI is starting to automate traceability and evidence generation, shifting validation from bottleneck to competitive advantage. Goldratt showed us decades ago: optimizing a non-bottleneck doesn't increase throughput. In software terms, faster coding without faster compliance just piles up inventory—more code waiting in the verification queue. But when you aim AI at the actual constraint—automating traceability, generating compliance evidence, orchestrating verification workflows—something different happens. You don't just boost productivity. You unlock the entire system. I first saw this pattern 25 years ago as a Rational Software alum working with iterative development. The goal wasn't writing code faster. It was lowering the cost of each trial. Lower cost of experimentation → lower risk → faster innovation. AI is now doing the same for regulated industries—not by writing code, but by proving it. Excited to see this in action: Next month (April 22–23), at COVESA's All Member Meeting in Porto, Portugal (Hilton Porto Gaia), we'll be demonstrating AI-powered approaches to these exact constraints—automating traceability and compliance evidence for automotive SDV workflows (ISO 26262-aligned). Looking forward to the Demo Showcase & Reception on the 22nd and discussions on unlocking verification bottlenecks in connected vehicles. If you're in the space, registration is open: https://lnkd.in/eAczs886 Where are you seeing the real bottlenecks in your regulated software releases? Coding velocity… or compliance & verification? Curious to hear your thoughts. #AI #RegulatedIndustries #SoftwareCompliance #DevOps #TheoryOfConstraints #ISO26262 #DO178C #IEC62304 #SDV #ConnectedVehicles #COVESA

18

1 comentário

If you're viewing this and wondering: "Why on Earth is Ben posting a random video of some Google folks talking about Cloudrun...IN PORTUGUESE?"...🤷‍♂️ Well...I just exported this video directly from Braiv to Linkedin using our upcoming "No-code Export" feature. What does this mean? It means using Braiv Export integrated with our favorite No-Code platfrom to auto-post a translated video to Linkedin!!! Watch this space #BraivOnes!!!

16

Some AI-exposed roles in Brazil are already down 5.4% in hiring Here’s the snapshot across the 10 most AI-exposed occupations • Falling: Debt-collection agents with -22% and Assistents with -5% • Flat: Receptionists, Office assistants and Accountants • Rising (for now): HR analysts with +9%, Outbound telemarketing agents with +8% and Systems analysts with +3% Read the full analysis → English: https://lnkd.in/dFJnVjp3 Português: → Portuguese: https://lnkd.in/duRFQD6j Thanks / inspiration • Erik Brynjolfsson, Bharat Chandar, Ruyu Chen — Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of AI • Tyna Eloundou, Sam Manning, Pamela Mishkin, Daniel Rock— GPTs are GPTs: Labor market impact potential of LLMs • Peter McCrory Alex Tamkin for AI Economic Index — (augmentation vs. delegation insights across occupations)

81

6 comentários

When we started Infrasity, we weren’t aiming for scale. We just wanted to help infra and AI startups explain what they’ve built in a way that actually makes sense to engineers. In less than a year, Infrasity has been working with engineering teams across the US, EU, and Israel, supporting devtool and infra startups doing up to $30M in ARR.  -> 80% of our customers retain  -> Most turn into long-term DevRel partnerships  -> And we ship at the speed of the engineering team: onboarding guides, working SDK examples, explainer videos, and more No paid marketing. No cold calls.  It’s been pure organic from day one & building in public, doing podcasts with B2B SaaS founders, staying close to our customers, showing up where they hang out at conferences, in Discords, reddit or deep in the web. Just pure product understanding + fast, technical execution. We’ve become the go-to content + DevRel engine for developer-first companies, shipping onboarding guides, SDK recipes, demo libraries, and launch content that actually helps teams go to market faster. All of this occurred within a year.  And we’re just getting started. Full story here: 👉https://lnkd.in/gAau6KBg #DevRel #DeveloperMarketing #SaaS Infrasity

8

Portugal’s Innovation Agency once told us we weren't doing R&D. Why? Because we weren’t inventing shiny new tech. Because software, they said, “isn’t innovation.” But here’s what we were actually doing: 👉 Working shoulder-to-shoulder with cannabis producers 👉 Testing software in real grow rooms 👉 Solving compliance pain on the production floor Their verdict? Just “gathering requirements.” Meanwhile, companies got their "R&D seal" status for… migrating to cloud servers. This mindset is stuck in the past. It punishes real-world builders and rewards checkbox projects that never leave the lab. At Cannabud.ai, we started with questions, not code. We went into licensed facilities. We listened. We mapped pain points. We turned those into hypotheses, tested them, and refined them under regulatory fire. While bootstrapping to fund it all. The result wasn’t just a platform. It’s an operating system for cannabis, built with growers, designed for uncertainty, and hardened in the real world. Our mission? Not flashy tech. But better days at work for cultivators: 👉 Clearer tasks 👉 Fewer compliance gaps 👉 No lost time, no paper trails 👉 Just control, traceability, and confidence Innovation isn’t about who’s first. It’s about who redefines the experience. That’s what Google did. That’s what we’re doing in cannabis. If building probably the most flexible, auditable, and user-driven ERP in Europe’s cannabis sector doesn’t count as R&D… Maybe the problem isn’t our definition of innovation. It’s theirs.

31

2 comentários

“software will be worthless” - AI companies need to focus on data. Larry Li , Partner at Amino capital - #websummit Lisbon. AI companies should be focused on building in a niche as more popular features will likely become features of larger foundational model platforms (aka Open AI, Anthropic etc)

7

Turning “no-shows” into new opportunities. 🚀 What happens when empty spots at the world's premier tech conference sparks inspiration in a future talent? Our partners at Startup Portugal gifted their vibe through TIMEPLACEME - and the magic happened instantly. Meet Anastasiya Tyurina - founder of Garage Tech Academy in Barcelona and one of TIMEPLACEME’s very first early adopters. Today, thanks to this initiative, one of her top students, Yaroslav, stepped into the world of 4YFN for the very first time. Instead of unused tickets - a magical "first impression." Instead of empty spots - a transformative experience that could define a career. We are thrilled to have gifted these moments. The magic worked - which means everything is going according to plan. After all, TIMEPLACEME was built exactly for these kinds of encounters to happen. ✨ #TIMEPLACEME #Lisbon #StartupPortugal #4YFN26 #Barcelona #TechAcademy #SerendipityAI #brave4real #ExperienceTech

15

2 comentários

Last week I talked about how kickoffs become theatre by April  https://lnkd.in/d83Wh4cd. A bunch of you asked: okay, so what actually prevents that? I've been thinking about this since I posted. I've implemented playbooks that have stuck and I've seen plenty that didn't, and when I look at what made the difference, it comes down to four things that most companies just don't do after the kickoff ends. 1️⃣ No operating rhythm.  2️⃣ Managers who can't coach what they're asking reps to execute.  3️⃣ Systems that don't reflect the new process.  4️⃣ And no way to measure if behavior is actually changing before you're already behind on the number. I put together a breakdown of what each of these look like and what you actually need to do to fix them. No, this isn’t theory - it’s things I've done that have worked. Swipe through if you're trying to figure out why your last rollout didn't stick. Or if you're planning one now and want to avoid the usual April fade. Which of these is missing in your organization right now? Drop a comment or send me a message - I'm curious what's the biggest gap for most teams. And if you want to go deeper on building this kind of sustained execution, EliteGTM and revlogic have partnered up and can help you. One example is the Academy which we built together.  It's not event-based training, it's ongoing enablement that keeps the playbook alive all year. 👉 Learn more: DM me, Oliver, Daragh or Mats

54

1 comentário

Most founders get MVPs wrong. They treat them like mini-products. Something that needs polishing, features, 'readiness'. But as Rafaela Azevedo (CEO & Founder, The Chaincademy) reminded our Copilot Group this week: Your MVP isn’t the product. It’s the proof. It’s the fastest, cheapest way to learn: 👍 What customers actually care about 👍 Which assumptions are true 👍 What’s worth building next 👍 Where the real risk is 👍 Not what looks impressive. A few takeaways Rafaela dropped that really landed 💡 POC vs MVP - Use a POC when you don’t know if something works. - Use an MVP when you need to learn with users. 💡 You only need 3–10 users to test - Quality feedback beats a crowd. 💡 After the MVP Analyse - iterate - refine - then scale. 💡 MVP trade-offs Fast and cheap, but limited, unscalable, imperfect by design. She also gave the group a brilliant crash course in prompt engineering (think better prompts - better prototypes - faster learning). Huge thank you to Rafaela Azevedo for bringing so much clarity and practicality to our Copilot Group session 🙌. If you would like to come join us on a future Copilot Group session join the wait list here https://tally.so/r/npaAPq or find out more here https://shorturl.at/fZds4

19

3 comentários

Deepak Jha 🤓 thank you for sharing this. It genuinely makes me happy to see the process land the way it’s supposed to. A little behind the scenes: I used to think I was good at “picking winners” when I helped run startup competitions… and my favorites almost never won. That was the moment I decided to stop relying on gut and start relying on something repeatable. Venture IQ is basically my attempt to translate Bill Reichert ’s “Getting to Wow” brain into a technical, auditable system. It reads a deck, scores it against a consistent rubric, and then explains the score with evidence. Not just “nice story / not for us,” but what is missing, where investors will push back, and what would de-risk it fastest. Here’s what it actually does in plain language: It generates an investability score with a dimension-by-dimension breakdown, so you can see exactly where you’re strong and where you’re exposed. It produces a traffic-light risk matrix that labels the real diligence risks (product, market, GTM, defensibility, execution, financing) and tells you what proof points would turn a yellow into green. It maps your deck to the questions investors and judges reliably ask, so you can rehearse answers before you walk into the room. And it gives slide-level guidance and copy upgrades so the story becomes tighter, more fundable, and closer to “Getting to Wow” standards. The original spark came from an idea by luis felipe valdes rueda: what if we built a startup judge that was less biased and less dependent on who you know? That’s what we’ve been building for a while now, and your post is a perfect example of why it matters. If anyone here wants the same kind of feedback, DM me your pitch deck PDF. I’ll run it through the system and send back the score, the risk matrix, and the highest-impact changes to improve it quickly.

8

If coffee is a fuel of capitalism, espresso is capitalism engineering its own fuel source. Yes, espresso is an engineering response to urban capitalism. That's why its history is unusually crisp and traceable. 2016 Portugal, I'd head to a café every afternoon with colleagues for "um bica"—a shot of espresso. Standing at the counter for three minutes with a tiny cup was the only socially acceptable way to pause between Wolfram Mathematica and long discussions. The coffee was the excuse. The ritual was survival. Coffee arrived in Europe in the 1600s. Early brewing was slow, communal, contemplative. Cafés were salons where you'd waste afternoons arguing politics. Then the Industrial Revolution happened. Cities densified. Lunch breaks shrank. Coffee needed to be fast. Get a hit, and off you go to work. In 1901, Luigi Bezzera patents a machine that forces hot water through coffee under pressure. Single cup. Seconds, not minutes. The word "espresso" captures it: expressly made for you, pressed under pressure, express speed. Early espresso was terrible—bitter, harsh, steam-scorched. Nobody cared. Speed mattered more than taste. You got the hit. The real breakthrough came in 1947. Achille Gaggia's lever machine hit 9 bar pressure. Two unintended consequences: better extraction and a golden foam on top. Customers thought the foam was a defect. The marketeer in Gaggia pushed it as "crema naturale." That foam became proof of quality. Espresso went from functional to sensorial. By 1961, electric pumps made it repeatable. Chains industrialized it. Italy kept it short, bitter, cheap, fast. The rest of the world stretched it into milk drinks. The Third Wave brought lighter roasts and single-origin obsessions. Some Italians say that's not espresso anymore. They're right. But the physics remained: ~9 bar pressure, ~25 seconds. Applied thermodynamics shaped by capitalism. We in India went from Nescafé sachets to third-wave coffee bars in fifteen years. I still don't see enough people shooting espressos. Too bitter, too intense. We'd rather douse ourselves in filter coffee with milk and sugar. Which is a funky proxy though, to check how deep rooted capitalism is in any given culture we're looking at. We're def not there yet. I'd love to meet all my fellow espresso lovers some day, but ain't nobody gonna travel just for a 3 minute shot eh?

34

4 comentários