Creandum✓

RT @Kathi_Wilhelm: Physical vibe coding is here. @Machenta_ brings instant learning to robots: show a robot a task once and it gets to work. You prompt it with a demonstration the way you prompt ChatGPT with text. While most of robotics races to collect more real-world data, @Ed__Johns and @vitalisvos19 spent the past years at @imperialcollege on the opposite bet: a model trained entirely in simulation that learns how to learn. That means it’s bounded by compute, not data, and it will only get better as it scales. We believe the hardest problems are often solved by people who bring a different lens. Ed is a true polymath: after his PhD, he spent a year as a singer-songwriter, playing gigs with a then still unknown Ed Sheeran. In a band, when someone shows you a chord, you rarely need to see it twice. Perhaps that’s why Ed asked a different question from most of his field: not how to make robots learn more, but how humans learn so fast. At @creandum, we're proud to co-lead Machenta’s $40M seed with @usv, backing Ed, Vitalis and the team.

RT @Ed__Johns: Today we’re out of stealth, and we’ve raised a $40m seed round co-led by @creandum and @usv, to bring Instant Robot Learning to life. For the past few months, we‘ve been busy building a new company. Machenta. This follows results from a new kind of robot foundation model we’ve built, which I’m so excited to share with you today! Usually, it takes hours/days of data collection and training to teach a robot. We’ve built a model that learns tasks instantly: one demonstration, no further training. A model trained entirely in simulation, without a single pixel of real-world data. A model where teaching robots is so quick and easy, it feels like physical vibe coding. A model for robots that learn like humans do, from someone like you. Welcome to the era of Instant Robot Learning. Thank you @Kathi_Wilhelm, @rebeccakaden, @hkanji, @jonno_evans and all of our angels for your support! @Machenta_ Learn more: See below for what you can do with Instant Robot Learning!

RT @ralfgulde: When we started @SereactAI , people told us returns were impossible to automate. This is Cortex handling @Zalando returns at a @cevalogistics fulfillment center. It’s live and in production, every shift. Proud of the team that made this real.

RT @_fredrik: nothing short of incredible continued growth displayed by @bernhardsson and team @modal. so very impressive. massive congrats on this double milestone very proud for @creandum to be a part of this journey since first round (led by @AmplifyPartners & @sarahcat21 - stellar job)

RT @Kathi_Wilhelm: 3) It's becoming table stakes for application companies to build their own models – both for defense (cost and access to models) and offense (better products) (3/5)

RT @MattJonWilson: Today we’re announcing our $40M Series A, led by @AirStreet, with participation from @MadronaVentures and @Creandum. Most of us leave one of the biggest decisions of our lives almost entirely to chance: how we spend our time at work. 🧵

RT @AgustinLebron3: I haven't said much recently about what I'm working on, but since this article is out now, I might as well say a few things. ➡️

The DeepMind trio who built a poker AI, are now making money for quant hedge funds

RT @Lifrordi: We are looking for more amazing researchers and engineers to join our team. EquiLibre Technologies is a frontier AI trading research lab. We train our own foundational models, use our own RL algorithms, and trade our own money. The proof is in the pudding. [THREAD]

RT @_ghazwa_: 14/ The best thing about us @creandum isn't the portfolio or the brand - but that we actually LOVE each other. Founders feel it the second they walk in the room. Grateful, thrilled, and humbled by our team (and big kudos to events!!!!!!) and blessed to represent our presence in the US💚

RT @_ghazwa_: 13/ to close it out, we got some vinyls involved and hosted all the @creandum family + our favorite ecosystem people at Charmaine's for drinks

RT @_ghazwa_: 12/ beyond the talks: yoga, kayaking, running, biking, duffy boats, sauna, Sausalito strolls. the summit was the scaffolding. the relationships are the point!

RT @_ghazwa_: 11/ alongside the mainstage, our very own @HanelBaveja broke out our founders into small rooms of ~12 founders, hosted by some great @creandum friends. 1/ Matt Wilson (Jack & Jill) on building in a post-Opus 4.5 world 2/ @ypatil125 (@appliedcompute) on lessons from building Codex 3/ @andrewdsouza (@boardyai) on running 1,000 FTEs vs 25 4/ @JULIOMARTINEZG (@Abacum_ai ) on cracking US GTM 5/ @DavideGrieco (@clay_gtm) on AI-native growth teams 6/ Wouter Van Den Brande (@DecagonAI) on AI-native GTM 7/ @lindh (Kive) on productivity per employee 8/ @MadhavanSF on monetizing AI products 9/ @TajTalks (@CoreWeave) on enterprise AI

RT @_ghazwa_: 10/ @lukeharries (head of growth @ElevenLabs) closed us out - by far the most tactical session of the summit Luke is one of those rare growth leaders who's gone from 0 → 1 → 100 in record time. Took ElevenLabs from $2m → $500m ARR. now the company is valued at $11bn. Luke had no growth background when he started. 3 lessons: 1/ Launch and launch again. New models = new capabilities = new viral moments (even if it’s unlocking European founders default to underselling - push every T3 launch toward T2. his formula in the picture 2/ Build growth systems, not campaigns. Litmus tests: if you go on holiday, does your channel still work? in the AI era, does it keep improving without you? 3/ Enablement engineering is the unlock. Give marketers @Lovable and @n8n_io, but gate it with a small team of friendly engineers who handle security reviews and approvals. Otherwise you get non engineers pushing untested code to @salesforce Advice for the europeans in the room: "Be a little louder"

RT @_ghazwa_: 9/ our very own @creandum VP @filiphmark moderated a design panel with @felixhaas (Lovable) and @iansilber (head of design @ OpenAI). A few takeaways: 1/@OpenAI's design lead said generative UI isn't ready. "nothing is truly generated on the fly. it's too slow, too unpredictable." imagine ChatGPT looking different every time you opened it. nobody wants that. the unlock is personalized but predictable - Spotify-style, where content adapts but the UI stays the same 2/ @Lovable users have stopped browsing the App Store. they just build the app themselves - finance, scheduling, comms, all in one place. the middle layer is collapsing in real time Felix: "building a product today is incredibly easy. building something people fall in love with is harder”

RT @_ghazwa_: 8/ on day 2, our very own @fredrik_ opened the floor to our portco founder @alanchang (@fuseenergy) hosted by @collision (@stripe) Couldn't have set this up better - @fuseenergy was at a few millions in ARR not that long ago and now a $5bn company. @stripe processes ~2% of global GDP and is up ~100x in volume over the last decade. 3 insights from John: 1/ Good company management is overdetermined. There aren't 1000 right ways - there's basically one: customer obsession, pushed against organizational entropy 2/ Stripe's agentic commerce bet starts with developers, not consumers. Vibe-coded apps today can do almost everything except the last step: registering the domain, paying for hosting, calling a real service. that's where Stripe is unwinding 20 years of anti-bot infrastructure to let Codex actually transact. the downstream is the real story: agents are anti-aggregator. they surface niche brands that couldn't reach the top of Amazon or Google. great news for the long tail, less great for the existing search layer 3/ Stablecoins are no longer a 2030 thesis - ~25% of Stripe's payment volume now flows through Bridge On building startups: “Markets are fair. If you deliver for customers, make them happy, follow through on your promises - that mode of happy customers compounds”

RT @_ghazwa_: 7/ Beautiful dinner at Shack15 with the SF skyline and golden gate bridge as the backdrop. Had very own @johanbrenner and @Kathi_Wilhelm give a heartfelt speech. Got @peteflint and our very own data scientist in resident Gustav involved to chat about network effects and data infrastructure in the age of AI @fritjofsson @shelgesson

RT @_ghazwa_: 6/ We had @juliajchou from @AbridgeHQ showcase her operator discipline. Abridge serves 250+ health systems sitting on 2bn+ clinician conversations a year, solving the #1 driver of physician burnout in america. 3 insights from her: 1. Make AI access frictionless from day one. Every single member gets unlimited Claude, ChatGPT, Gemini, Cursor. the unlock is access itself 2. Enablement is a daily ritual, not a policy doc. The company pushes for AI demos at every all-hands and info sharing across internal slack channels (#ai-hype-machine, #ai-tools-for-all). They have 30 context agents fan out across RSS/HN/Reddit/X each morning to surface the top 5 AI posts relevant to their space 3. Before each hire, install an AI headcount test. Can AI close this gap in 6 months? Is this role trust-dependent or workflow-dependent? Are we hiring for today's ceiling or tomorrow's floor? Her line: "How you build is what you build"

RT @_ghazwa_: 5/ our very own @simonschmincke set us up for Tom Blomfield’s (@t_blom, YC, prev @ Monzo) talk, which landed personal for many entrepreneurs in the room. Tom built Monzo from "this'll never work" to 25% of UK adults having an account. Almost every investor on his cap table was American - and that’s not pure luck. A few things we learned: 1/ Tom showed us some sayings german and english people use to put ambitious people in their place “tall poppy”, “above your station”, “who does he think he is”, “the bigger they are the harder they fall” - Americans don't really have equivalents, the American default for ambition is "tell me more" 2/ Permission to dream is the SF unlock. "in the UK they say you're an idiot. here they say - how can I help" 3/ You don't have to relocate everything, but sales has to be here. Tom watched Monzo's American cap table form because american capital believed before british capital did Takeaway: Take more risk

RT @_ghazwa_: 4/ Next up, our @creandum founder @staffanhelgesson hosted @AppliedInt CEO Qasar Younis (@qasar). Qasar is the definition of intentionality. Every detail at @AppliedInt is engineered, from culture to counter-positioning. they now build for everything from Porsches to fighter jets. 3 insights from him: 1/ Culture is just the way two people behave with each other. Everything else is downstream - feedback, decisions, disagreements. the smallest-unit interactions, multiplied across the whole org 2/ Founder culture is not a values doc. Don’t wait for a scheduled all-hands to address culture problems - tackle issues head on immediately. 3/ Being an outsider, especially in Silicon Valley, can help because you’re less constrained by what everyone thinks is possible. We’re entering a period where small teams can do things that previously required enormous organizations, and that creates opportunities for founders who are willing to think in a new way.

RT @_ghazwa_: 3/ Our very own @nicckopp and @creandum gp @ @specht_p kicked us off on winning in the US Nic is the definition of relentlessness, and enabling CFOs to go for zero-day close at Rillet is the testament. 3 insights from him: 1/ Move where the gravity is, and move fast. After many redeyes between NY and SF, Nic decided to move his whole company and family within a few weeks. If you're truly serious about the US, you need to. you can't run it from the back seat. 2/ Ride the micro-hype cycles. Sometimes, being European defaults founders to undersell. Don’t hold back. When Opus 4.6 dropped (on a Saturday), Rillet had 1,000 people sign up for a "Vibe Coding for Finance" webinar by Tuesday. This has become a weekly live workshop 3/ The bar for early hires needs to be exceptionally high in a way most founders aren't willing to enforce. Rillet tests for scrappiness, drive to its mission, and deep AI curiosity, not pedigree