long hoang

Currently, rollups are fragmented, leading to poor UX and disjointed liquidity. Here the first demo of Javelin by our portco NodeKit With Javelin, users can synchronously send messages and move funds across #rollups on NodeKit within a single 2s block.

11

🇩🇪 Funding Rounds in Germany - November 4-8 🇩🇪 Last week saw activity across tech, telemedicine, and gaming. Here the latest funding rounds... 🧠 MONA AI - Recruitment Process Automation - Saarbrücken - €2M (Undisclosed) Lead Investors: Undisclosed 📡 Xavveo - Photonic Radar Sensors - Berlin - €7.9M (Seed) Lead Investors: imec.xpand; and Vsquared Ventures 🛠️ Dash0 - Developer Observability Platform - Solingen - €11.42M (Seed) Lead Investors: Accel 💻 TCC GmbH - Telemedicine Solutions - Hamburg - €20M (Series A) Lead Investor: Thomas Pötzsch 🎮 Gunzilla Games - Gaming Studio - Frankfurt - Undisclosed (Venture Round) Lead Investors: VanEck 🏭 ANTICIPATE - AI-Powered Manufacturing Video Analytics - Aachen - Undisclosed (Non-Equity Assistance) Lead Investors: BIND - Basque Open Innovation Platform Did I miss any? Drop them in the comments. As ever, congratulations to all involved. LP ✌️ Pack GTM | SaaS Sales Recruitment in Germany #sales #hiring #careers #startups #recruitment #venturecapital

22

4 Comments

🚨 Ignore This at Your Own Risk! 🚨 Web3 👇 Blog 👉 https://lnkd.in/gYfxv9rH 🌐 Discover the • Ins and outs of transaction • Execution on Layer 2 chains, • with a special focus on ZK Rollups 💯 👉 Save yourself from • The noise and • Hassle of sorting irrelevant Web3 content. 😎 #zkEVM #layer2 #zkRollups

5

3 Comments

Transformers, Tokenization, and Embeddings 🤯 Are you confused with LLM terms but at this point you are afraid to ask? No worries, you are not alone! Together, we’ll simplify these complex terms so you can confidently discuss and use them. ⚙️ 𝗧𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺𝗲𝗿 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 The backbone of modern LLMs, utilizing layers to process sequences in parallel for efficient learning and adaptation. 🔠 𝗧𝗼𝗸𝗲𝗻𝗶𝘇𝗮𝘁𝗶𝗼𝗻 Converts text into smaller units (tokens) for processing, crucial for understanding and managing linguistic data effectively. 🔧 𝗙𝗶𝗻𝗲-𝗧𝘂𝗻𝗶𝗻𝗴 Refines a pre-trained model to enhance its performance on specific tasks or datasets by adjusting parameters. 🎓 𝗣𝗿𝗲-𝘁𝗿𝗮𝗶𝗻𝗶𝗻𝗴 The process of training a model on vast amounts of data before it is fine-tuned to specific tasks. 💠 𝗘𝗺𝗯𝗲𝗱𝗱𝗶𝗻𝗴𝘀 Transforms words into dense vectors that capture semantic meanings, enabling models to understand text similarities. 🎩 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗠𝗼𝗱𝗲𝗹𝘀 These models can autonomously create text, simulating human-like writing or responses based on training data. 🔎 𝗗𝗲𝗰𝗼𝗱𝗶𝗻𝗴 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀 Techniques used to select the next word in a sequence when generating text, crucial for coherent outputs. 👁️ 𝗔𝘁𝘁𝗲𝗻𝘁𝗶𝗼𝗻 𝗠𝗲𝗰𝗵𝗮𝗻𝗶𝘀𝗺𝘀 Focuses on different parts of the input data, determining which aspects are most relevant for a task. 💬 𝗟𝗟𝗠 𝗣𝗿𝗼𝗺𝗽𝘁𝗶𝗻𝗴 Employs specific cues or prompts to guide the AI’s output, enhancing the relevance and accuracy of its responses. Got questions or want more details?  Feel free to ask in the comments below. Also, let me know what LLM topics you're curious about—we could explore them in future posts! ♻️ Share this post and teach your network something new.

62

11 Comments

Transformers, Tokenization, and Embeddings 🤯 Are you confused with LLM terms but at this point you are afraid to ask? No worries, you are not alone! Together, we’ll simplify these complex terms so you can confidently discuss and use them. ⚙️ 𝗧𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺𝗲𝗿 𝗔𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲 The backbone of modern LLMs, utilizing layers to process sequences in parallel for efficient learning and adaptation. 🔠 𝗧𝗼𝗸𝗲𝗻𝗶𝘇𝗮𝘁𝗶𝗼𝗻 Converts text into smaller units (tokens) for processing, crucial for understanding and managing linguistic data effectively. 🔧 𝗙𝗶𝗻𝗲-𝗧𝘂𝗻𝗶𝗻𝗴 Refines a pre-trained model to enhance its performance on specific tasks or datasets by adjusting parameters. 🎓 𝗣𝗿𝗲-𝘁𝗿𝗮𝗶𝗻𝗶𝗻𝗴 The process of training a model on vast amounts of data before it is fine-tuned to specific tasks. 💠 𝗘𝗺𝗯𝗲𝗱𝗱𝗶𝗻𝗴𝘀 Transforms words into dense vectors that capture semantic meanings, enabling models to understand text similarities. 🎩 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗠𝗼𝗱𝗲𝗹𝘀 These models can autonomously create text, simulating human-like writing or responses based on training data. 🔎 𝗗𝗲𝗰𝗼𝗱𝗶𝗻𝗴 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀 Techniques used to select the next word in a sequence when generating text, crucial for coherent outputs. 👁️ 𝗔𝘁𝘁𝗲𝗻𝘁𝗶𝗼𝗻 𝗠𝗲𝗰𝗵𝗮𝗻𝗶𝘀𝗺𝘀 Focuses on different parts of the input data, determining which aspects are most relevant for a task. 💬 𝗟𝗟𝗠 𝗣𝗿𝗼𝗺𝗽𝘁𝗶𝗻𝗴 Employs specific cues or prompts to guide the AI’s output, enhancing the relevance and accuracy of its responses. Got questions or want more details?  Feel free to ask in the comments below. Also, let me know what LLM topics you're curious about—we could explore them in future posts! --------- ➕ Follow @Sebastian Illing for the latest AI news. ♻️ Repost to keep your network up to date.

397

6 Comments

List of Startups Receiving Funding in #Germany. Check out the list of #Germany-funded startups available for your reference at any time. View the list now https://lnkd.in/dpdZUJyi #startup #startups #vc #venturecapital #investmentbanking #privateequity #managementconsulting #b2b #founders #entrepreneurs

1

TON Foundation X Curve Finance In an exciting collaboration, the TON Foundation and Curve Finance are joining forces to launch a stable swap project on the TON blockchain. With stablecoin adoption on the rise—USDT on TON hitting $700M in circulation within just four months—this partnership aims to make stablecoin trading more efficient and user-friendly on the TON network. Notable features of this partnership: - Curve Finance's stable swap formula will be integrated into TON, creating a robust trading infrastructure. - veCRV holders and other qualified users will receive airdrop rewards from the project’s native tokens. While TON is expanding, it faces challenges in DeFi, including a limited lending ecosystem. However, with this new collaboration, TON is positioning itself to become a key player in the global DeFi space. 💡 My take: With this partnership, TON is making bold moves to boost its DeFi capabilities, and the involvement of Curve Finance could be the game-changer TON needs to stay competitive. #defi #blockchain #web3 #decentralization

18

7 Comments

As Altura gears up to further scale our Engineering organisation, I sought valuable input from someone who’s been through it: Daniel Gebler, CTO at Picnic. Here are key insights from Daniel: At Altura, we implemented Shape Up to structure our Product & Engineering process (more on that later), but my chat with Daniel highlighted several critical elements for success: 𝟭. 𝗙𝗼𝘀𝘁𝗲𝗿𝗶𝗻𝗴 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝘆 𝘄𝗶𝘁𝗵 𝗔𝗹𝗶𝗴𝗻𝗺𝗲𝗻𝘁: Scaling means teams need independence without losing sight of the company’s objectives. Daniel emphasised the importance of well-defined domains so each team can act autonomously yet remain connected, helping us grow in sync without friction. 𝟮. 𝗜𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗶𝗻𝗴 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗲𝘀 𝘁𝗼 𝗔𝘃𝗼𝗶𝗱 𝗖𝗵𝗮𝗼𝘀: Growth without structure often leads to inefficiency. We discussed how adaptable yet disciplined frameworks prevent that. Process matters—it turns growth into progress rather than just activity. 𝟯. 𝗠𝗮𝗶𝗻𝘁𝗮𝗶𝗻𝗶𝗻𝗴 𝗮 𝗦𝘂𝘀𝘁𝗮𝗶𝗻𝗮𝗯𝗹𝗲 𝗚𝗿𝗼𝘄𝘁𝗵 𝗣𝗮𝗰𝗲: Scaling isn’t just about speed and balancing velocity with a lasting foundation. Daniel shared why it’s essential to occasionally slow down and strengthen foundational elements, ensuring we’re built to last. Thanks to Daniel for his insights on scaling thoughtfully and sustainably. I am very excited to see the increased innovativeness and velocity that scaling will bring 🚀 🚀 P.S. Every day they provide a great lunch for everyone, how nice is that! We need a bigger kitchen ;)

45

2 Comments

We need to start re-thinking the idea of no-code. It's a naming convention built by those who can code. When the car was built, it wasn't named no-horse. Whop represents the car in the innovation evolution. Maybe no-code should just be redefined as Whop.

14

1 Comment

📣 Free founders-only event below! I often see founders get a lot of early traction in signing customers and then hit a wall when trying to scale that repeatedly. Why?? Maybe it's because: 1️⃣ that 1st group of customers came through warm intros from your accelerator batch 2️⃣ you've already combed through industry connections from your previous career 3️⃣ you're just good at brute-force selling but you're only one person 4️⃣ you are still narrowing in on product market fit Next week I'm hosting a free, founders-only candid discussion with two people who have had to break through this wall out of necessity and now coach countless founders on this. More on our speakers: 📈 Ian Illig took his 1st job out of college at a friend's startup (employee #1). He quickly found himself trying to sell a product that was still finding its market fit. Fast forward years later and Ian had completely built out the sales team and revenue, allowing the company to get acquired. And then he turned around and did the same thing again with the same founding team. 👀 Sonia Baschez was one of the 1st employees at Atmos, a consumer & builder-facing product for homebuilding. She developed early marketing channels & helped the company grow from Pre-seed to Series A. Sonia optimized product launches to reach the right customers, a challenging feat given their core customers are actively NOT in the tech bubble (aka Techcrunch and a thread on X are not going to do the trick). Whether they know it or not, both Ian and Sonia have been secret superpowers for me as I've grown The Council Fund and supported portfolio companies. I'm excited to share their playbooks & expertise with founders for FREE next Wednesday. Join us 12-1pm Pacific! Link to RSVP in comments below ;)

52

3 Comments

After 1000+ user conversations over the past 3 years, 3 pivots, and analyzing signals from similar startups, Aaron Torres and I decided, after a bunch of consultations, to sunset Klotho and return the remaining funds to our investors. It came down to believing there's no market for the problem we're trying to solve that sustains a startup, and that the closest problems we could solve with a reasonable degree of risk would require a harder reset in both team and tech which means limited time. I still firmly believe in the vision - it was beautiful seeing early customers and users realize parts of it when using Klotho and InfraCopilot. I'll be cheering from the sidelines for folks looking for a way to get us closer to it. It's never a happy moment to stop a dream before it's realized, but it feels like the right thing to do. A big thank you to believers! The team, users, customers, friends and investors - everyone risks something when believing in what could be next. We are forever grateful! If you want more details, feel free to reach out - I'm happy to share more. I'll be taking a stretch of time off (first time since 2006), reading books, talking to friends, and thinking about what I'd like to do next. So if you need an excuse to re-connect, this is a wonderful opportunity <3 DMs are welcome!

254

39 Comments

Some of the biggest differences I've noticed on the engineering side working in a start-up like environment within Google X vs building a product from scratch at DealPage: 1. Infrastructure: at X, everything ran on Google infra. It was extremely reliable, but there was a learning curve to really understand how jobs were scheduled and executed. At DealPage, we rely on various infra providers which abstract away a lot of the complexity. 2. Buy vs Build: Almost everything at Google X is built in house or by Google. At DealPage, we look for open source solutions as often as we can, and resort to buying only when no good open source option exists and we don't have the capabilities to build it ourselves. 3. Execution Speed: I thought I was moving fast at X with a small team. Building a product from 0 and launching last year during YC was a different story, especially when your team is just two people. Curious if anyone else has had a similar experience?

28

1 Comment

From zero to hero in 6 months? 🚀 I thought it was impossible. But that's exactly what happened when a skeptical startup founder finally gave me a chance. Here's how it went down: • Company: E-Commerce SaaS startup • Challenge: Scaling team from 10 to 50 • Timeline: 6 months • Initial stance: "Headhunters are too expensive" The founder's first words to me: "Prove me wrong." Challenge accepted. Month 1-2: • Filled 5 key leadership roles • Reduced time-to-hire by 40% • Quality of candidates? Off the charts Month 3-4: • 15 more hires across engineering and product • Diversity increased by 35% • Employee retention rate: 95% Month 5-6: • Final 20 roles filled • Scaling goals met 2 months early • ROI on headhunting fees: 300% The founder's words now? "I can't believe we waited this long. You've transformed our business." From skeptic to believer in 180 days. This isn't just a success story. It's a wake-up call. Are you still on the fence about headhunters? Let's chat. Your success story could be next. Drop a 👋 in the comments if you want to learn more about scaling your startup the smart way. #StartupGrowth #HeadhuntingWorks #ScalingSuccess

21

7 Comments

Incredible: StackBlitz's Bolt hits $4M ARR in just 4 weeks! Imagine describing your dream web app in plain English and watching it come to life in minutes. That's bolt.new. No coding is required. Powered by Claude 3.5, it handles everything from frontend to deployment - all in your browser. The numbers tell the story: From $0 to $4M ARR in 4 weeks, with a 99% reduction in development costs. One user turned a $5K, 3-month Upwork quote into a 2-week build for just $50/month. You might still hit some bugs, but the team usually fixes them quickly. Got an app idea you've been sitting on? Time to build it! 💡

18

3 Comments

Many developers face challenges when testing decentralised exchange (DEX) solutions on testnets, especially due to Uniswap testnet's unavailability. To address this, I developed the Web3Swag Testnet Dex specifically for the Sepolia & Holesky testnets. Now, developers can seamlessly test their DEX solutions on these alternative test environments. Tryout: https://lnkd.in/gM_GeR2z #Blockchain #BlockchainDevelopment #Web3 #Web3Development

14

Funding Secured for Cloo AI! 🚀 I’ve devoted 20 years of my life to mastering software engineering. Software is my bread and butter, my complicated relationship with many ups and downs, and my biggest passion. Starting with POS software for selling Schnitzels, I’ve worked my way up to large enterprises and Fintechs in Switzerland. Along the way, I’ve learned 7 programming languages and ventured into the labyrinths of AWS, GCP, and Kubernetes, emerging undeterred. I’ve also embraced leadership, building the team that scaled Batmaid, a major international player revolutionizing the home cleaning industry and improving the lives of hundreds of thousands of people. Now, I’m embarking on my next adventure, starting from scratch again. I believe that the software industry is in trouble. We’re overwhelmed and oversaturated with software products. Every business relies on software to grow, yet finding the right tools and making the most out of them is an enormous pain, for small and large companies alike. The internet is saturated with outdated promo sites and SEO spam. Making an informed decision is overwhelmingly difficult. SaaS vendors spend up to 40% of their resources on finding the right customers – money that could be better spent polishing their products and listening to clients. Yet, the SaaS industry is expected to triple in size by the end of the decade. Like energy consumption, the capacity for information processing is a fundamental aspect of life. It will only go up. Buyers want quality information and intelligence. Hence: buyer intelligence, the next big thing. This is why I’ve decided to devote my life to this problem by co-founding Cloo AI. Together with my co-founder Michal Kaczor, we’re going to stubbornly work on this problem so that buying software becomes as easy as ordering an Uber. (Remember ordering taxis pre-Uber? That’s how it feels today to buy business software.) We’re building a magical tool that leaves you informed, at ease, and better prepared to solve the problems that really matter to your business. We’re very fortunate to be generously backed by 🚀Michal Rokosz from Inovo.vc, Poland’s top venture capital, as well as an incredibly successful and talented group of Angels, including Bartek Pucek, Charles ▫️ Delingpole, Przemek Gotfryd, Miguel Fernandez Larrea, Jan Domanski, PhD . Thank you for believing in us and our mission! If you’re adventurous, passionate about LLMs applied to software, and eager to explore uncharted territories in engineering, join our highly energetic and fun-loving team. Early employees receive equity, which multiplies as the business grows. We’re recruiting: Full Stack Engineer, Warsaw (https://lnkd.in/dz5zQmif) Growth Associate, Warsaw (https://lnkd.in/dqrhfAs7) Check out our job postings and become part of something transformative!

101

17 Comments

When people ask me how Web3 is re-defining ownership, I often do not go straight to topics like real-world asset tokenization - I leave that to my colleague Dr. Matthias Reiser 😉. I go for something with which many have ample experience: gaming.   NFT games are at the forefront of a true transformation in gaming and ownership.   Creating and playing an NFT game in a digital economy will bridge virtual and real worlds by adding a layer of ‘real-life’ NFTs, thus stimulating and rewarding player engagement.   What’s more, the decentralized character of Web3 games, powered by blockchain technology and NFTs, ensures that in-game items and characters are not controlled by a central authority but distributed across the blockchain.   With every transaction and ownership transfer being recorded on the blockchain, manipulation or fraud within the ecosystem will become negligible.  Talking about ecosystems, NFTs provide interoperability to Web3 games, allowing players to use their digital assets across different gaming platforms. And so the fusion of NFTs and gaming will allow players to engage with blockchain-powered technology, thus becoming architects of a decentralized gaming ecosystem while increasing their ability to earn.   Yet no matter what the application, it’s important to note that data for NFT images are not stored on the blockchain; rather, they are often web links that can very easily be deleted. The implications of this are that if web links are indeed erased, proof of ownership is impossible to prove, placing NFT security as a key concern for industry.   If you want to have deep conversations about Web3 gaming, security, and blockchain, join me at Proof of Talk in Paris, 10-11 June. https://lnkd.in/eVHmwkne    #PoT24 #ProofOfTalk #Web3 #gaming #NFT 

22

1 Comment