Melbourne Meetup: Agentic Coding Trends 2026

Tonight is the night, hoping to see you there, whatever you level in AI, this Meet Ups are all about building the local community of people interested in all things generative AI, agents and builds. Three cool talks tonight - March 31, 2026 6:00 – 6:20 PM | Arrival, socialising & networking 6:20 – 6:40 PM | Community updates + latest GenAI & Agentic AI news 6:40 – 7:10 PM | 📊 Anthropic's 2026 Agentic Coding Trends Report - panel discussion. 7:10 – 7:25 PM | 📊 OpenClaw Walkthrough with Gaurav Caprihan, Ph.D. Learning about this new craze sweeping the agentic world. 7:25 – 7:45 PM | 🎵 Creative GenAI: Song to Music Video — Instructor-Led Walkthrough An end-to-end creative AI session: song idea → lyrics → full track → music video using modern generative AI tools, with the boss George Vossos 7:45 – 8:00 PM | Q&A + closing networking 8:00 PM | 🍻 Optional: head to a nearby pub for dinner Looking forward to this!

🏛️ The Digital Renaissance: Where History Meets the Future of Tech If we were building the web during the 15th century, what would our "stack" look like? I’ve been thinking a lot about the parallel between the Renaissance era and our current Generative AI revolution. Both periods are defined by an explosion of human creativity, collaborative research, and the birth of entirely new "tools" that redefine what’s possible. Imagine collaborating with a team of developers and research scientists, blending the artistic mastery of the masters with the computational power of today. 🎨💻 As an Infinite Networker, my goal is to maintain that same spirit of multi-disciplinary innovation: * Engineering with Precision: Just as architects of the past obsessed over symmetry, we focus on Core Principles and Model Architecture. * Artistic Vision: Using technology to push the boundaries of Creative Design and Content Creation. * Collaborative Research: Working across roles—from data scientists to developers—to ensure Global AI Adoption is grounded in quality. We aren't just coding; we are part of a modern-day workshop of ideas. The tools have changed, but the mission remains the same: Unleashing Software Advantage through curiosity and collaboration. How are you bringing a "Renaissance" mindset to your dev team this week? 🚀 #FullStackDeveloper #GraphicArtist #TechInnovation #GenerativeAI #SoftwareEngineering #DigitalRenaissance #DevCommunity #GitHubCommunity #AITrends

As part of our AI-first initiative, working with Claude code has been truly transformative. It hasn't just improved my output; it has fundamentally redefined my workflow. By shifting how I approach tasks, I’ve seen significant gains in both quality and productivity. Most impressively, the window between ideation and implementation has shrunk from days to hours, opening up possibilities that were previously out of reach. The key piece is to invest in learning about the tool and its eco-system.

Last night, a Trønder convinced me to stop typing. 🇳🇴 Let me explain. I went to my first Bergen meetup, Claude Code: Vibekoding i praksis, organized by Hubbel Kodeklubb. I went in thinking I had a decent handle on the tool. I've been building with Claude Code for weeks. PRDs drafted and pushed to Confluence. Clickable prototypes. Parallel workflows eating hours that used to fill days. But the presenters did something I wasn't doing: 🔊 They talked to the AI instead of typing. Here's what impressed me... The presenters were from Trondheim. If you know Norway, you know: that's not the easiest dialect to understand. And the app they used understood every word! Watching others use a tool you think you know is a mirror. It shows you what you're missing and what you've already got right. Three things I'm taking home: 🎙️ Voice-first input — thinking out loud beats typing 🔁 Watching others work is wildly underrated as a learning strategy 🤝 Bergen's tech community is vibrant... and I want to be part of it To the Hubbel kodeklubb organizers: thank you! For a newcomer in town, this was exactly the kind of room I needed. 🇨🇴🇳🇴🚀 👇 Are you using voice tools in your AI workflows? What's changed for you? #ClaudeCode #VibeCoding #Bergen #AIWorkflows #BuildInPublic

Stop pretending LLMs are just another API; they’re stakeholders with a mind of their own. Let’s be honest: we’ve all had enough of "AI will replace us" webinars. If you’re a Middle+ designer, PM, or founder, you know the real pain isn't about Midjourney prompts. It’s about building a product when the "engine" (LLM) is unpredictable and fundamentally changes how users interact with logic. LLMs are not neutral tools. They have patterns, biases, limitations, and behaviors that directly affect what your product can and cannot do. When you build on top of an LLM, you inherit its worldview. That makes it a stakeholder by definition. I’m hosting the first Projector Global Community meetup right here in the Buenos Aires tech scene. We’re skipping the basics to talk about the real shifts in our craft: - shifting from visual design to logic and intent-based design; - how AI capabilities (and limitations) should pivot your roadmap; - the role of a Designer in an AI-driven decision cycle. No boring slides. No buzzword bingo. Just 10-12 people, good coffee, and a deep dive into the messy reality of building AI products (and how to avoid the "AI Wrapper" trap). Details: 📅 April 2nd | 18:00 - 20:00 🌐 Language: English 🎟 Free, but spots are limited to keep the discussion high-quality. How to join: Registration via Luma is mandatory to keep the discussion high-quality and manageable. You’ll find all the meeting details and the exact location there. Link for registration in the first comment below! 👇 #BuenosAiresTech #BuenosAires

Most people are using Claude… But they’re only using one-third of it. Anthropic didn’t build one tool. They built three layers — and when combined, that’s where the real productivity unlock happens. 🖥 Claude Code → Build & debug inside your codebase 🤖 Claude Cowork → Automate docs, research & office workflows 📁 Claude Projects → Align teams with shared context The magic isn’t in choosing one. It’s in using all three together. Use Code to build it. Use Cowork to manage it. Use Projects to align the team on it. If you're serious about AI productivity in 2026, understanding this ecosystem is non-negotiable. 💬 Which one are you using right now — Code, Cowork, or Projects? #ClaudeAI #Anthropic #AIProductivity #AIAgents #FutureOfWork #DeveloperTools #AITools #TechLeadership

We hosted a 250+ person AI hackathon at WorkOS HQ in SF. One thing that stood out: The nature of hackathons is changing fast. You don’t need 48 hours anymore to build something real. With the right tools, teams were shipping fully functioning apps in hours — not days. What mattered wasn’t time. It was taste, clarity of idea, and how fast you could iterate. The best projects didn’t come from pre-formed teams. They came from: - random pairings - people who had never met before - builders who just decided to start There’s something about that environment - high trust, low friction, shared curiosity - that creates outsized outcomes. A few quick takeaways: - AI is compressing build time, but raising the bar on what “good” looks like - distribution + storytelling matter just as much as the product itself - hackathons are becoming more about collision than code - strangers are still the best co-founders (for a weekend) Recap below 👇

In video post-production, structure is everything. Named tracks, organized timelines, clean project folders. Thursday I felt that same satisfaction inside an AI workflow for the first time. Session 01 of the "Weavy Live Studio" was about exactly that. Not prompting harder, building smarter. Three things that stuck: - System Prompts are the brain. Define the role, the constraints, the output structure. Your LLM stops guessing and starts delivering. - Variables make everything modular. Swap a product, a scene, a style without rewriting from scratch. - Rename everything. Your node names become the labels your client sees when you deploy the workflow as an app. In the live session we built a branded cap with Weavy and Lighthouse AI Academy logos. I took that workflow and pushed it further, turning it into a basketball jersey. Same system, different output. That's the whole point. Five more weeks with Raoni Lima and the Weavy Live Studio. The system is just getting started. 🤘

“Svetlana, thank you so much! You always break the mold!”, – that’s what one of the founders said at my Weekly AI Mentoring session yesterday. First of all, it’s genuinely nice to hear :) A big-league guy running a million-dollar business finds something at every one of our sessions that saves him time. Second, let’s unpack what that actually means. Because “breaking the mold” in AI isn’t about magic: 1. The death of copy-paste and the shift to File System The typical non-tech workflow with AI looks like this: copy code, paste into chat, get answer, paste back. I showed how to connect Claude directly to a folder on your computer. The AI sees your entire file structure at once. Context never gets lost again. BOOM 2. AI can break out of the browser…with hundreds of ready-made MCPs And it’s not just for geeks anymore. We connected Claude directly to Figma and built a solid one-pager in 5 minutes. BOOM 3. Prompt engineering is deprecated 😄 Well, not exactly. But why obsess over wording when what actually matters is giving the system context and teaching it to ask you the right questions?! We built an Agent together in Claude Cowork mode: one that sets its own tasks, picks its own tools, and delivers a finished result. I’m against AI still staying in the hands of a few dozen percent of advanced users. And I will keep running my mentorship with the kind of people who can actually turn it into money🤩 Next sessions cover: automated scheduled tasks and building root project architectures. Keep growing. #AI #FoundersProductivity #FutureOfWork