Vibe Coding Products in Bangkok, Volume 1

◎ Online

Hosted by Sathpal S. / Green Software - Scotland

Compare local AI models with fast frontier systems in this practical Green AI talk, hosted online by Green Software – Scotland on 5 October 2026. Local AI or the frontier? Running a model locally can appear to be the greener option, while lightweight frontier models promise efficiency through speed and scale. But which approach actually has the lower energy use and carbon impact? This one-hour virtual session examines that question without relying on assumptions or vendor-led claims. Marcus Knight, founder of BeGreenAI, will put local models head to head with fast frontier options such as Claude Haiku and GPT-4.1 nano. He will explore where each approach wins, and why neither side offers an obvious universal answer. The session is designed to help attendees make better everyday AI choices. You can expect to learn: - How local and frontier models differ in their energy and carbon implications - When reaching for a local model may make sense - When a fast, lightweight frontier model may be the better choice - Why model efficiency should be judged through evidence rather than assumptions - How greener AI decisions can also improve workflows, productivity and results Speaker and format Marcus Knight is an AI trainer and sustainability consultant working at the intersection of AI adoption and environmental impact. Through workshops, talks and strategic support, he helps teams audit how they currently use AI, reduce digital waste, and adopt greener tools and systems. His approach focuses on practical, evidence-based sustainable innovation while helping organisations get more from the AI products they already use. Hosted by Sathpal S. and the Green Software – Scotland community, this autumn event takes place online from 5:00–6:00pm BST on Monday 5 October 2026. The virtual format makes it accessible to attendees beyond Edinburgh, with the event link becoming visible to registered participants. Meetup’s event chat also gives attendees a place to connect and continue the conversation.… Register for this event

ClaudeChatGPT

Sponsors: Green Software Foundation

⌖ New York

Hosted by tokens&

Build a production-ready real-time video AI agent in New York on 9 October 2026, with expert support, GPU compute and a ready-to-use builder stack. The VAST Builders Challenge is a one-day technical hack event for turning video archives into useful, searchable AI applications. Rather than starting from scratch, teams receive videos, infrastructure and development tools so they can focus on creating a shippable prototype. This New York edition forms part of a three-city series spanning New York, London and San Francisco. What you’ll build Choose your own video-agent concept and experiment with a stack covering: - Video understanding: reason over clips using NVIDIA Cosmos - Semantic search: find moments across hours of footage with natural-language queries - Object detection: run real-time detection and tracking across frames with YOLO - LLM-powered logic: use models from Weights & Biases to drive agent behaviour and application workflows The VAST AI OS handles video ingestion and the data layer, while models run on CoreWeave GPUs. Builders will ship using Cursor, with expert guidance and compute available throughout the day. Format and schedule Doors open at 8:30am EDT with breakfast, followed by opening keynotes at 9:00am. Building begins at 9:30am and continues through lunch before teams present demos at 4:30pm. The event concludes with awards at 6:30pm. Projects will be assessed by AI and infrastructure specialists from NVIDIA, VAST Data, CoreWeave and Weights & Biases, including Diego Garzon, Ravi Garg, Ram B., Brian Verkley, Prashanth Nalubandhu, Brandon Kates and Vera A.… Register for this event

Cursorn8n

Sponsors: VAST Data · CoreWeave +3

◎ Online

Hosted by The Curve and Homeslates

A focused virtual webinar for senior leaders moving an AI-built prototype into production, covering security, data handling, scale, ownership and trust. AI has made it dramatically easier to create working software, but a convincing demo is not the same as a system an organisation can safely rely on. Running from 11:00am to 11:30am on 13 October 2026, this session examines the work that emerges once a vibe-coded tool leaves the experimental stage, starts handling real data or becomes part of a customer journey. From working prototype to trusted system James Ridgway, CTO at The Curve, will be joined by Tom Gothard, Managing Director at Homeslates. Together, they will explore what changes when an AI-built application moves from “it works” to “we depend on it”. You can also view James Ridgway’s profile for more on his background. The discussion will address questions that are easy to postpone during rapid prototyping: Who owns the tool? Where does its data sit? What happens when it breaks? Can it operate securely and reliably at scale? The speakers will unpack the confidence gap between seeing a demo function and knowing that the underlying system is safe to implement. Topics include: - Why easier software creation does not automatically make software easier to trust - What a working demo cannot reveal about security, data handling and scalability - The implications of using a tool with real organisational or customer data - How to decide whether to review, harden or rebuild an existing application - Why platforms such as Lovable and Claude Code can help build software but cannot independently provide credible assurance that it is safe Who should attend? This webinar is designed for senior leaders across all industries who already have something built with AI and need to determine what happens next. It is not an introduction to vibe coding, and no technical background is required. Instead, expect a concise, practical conversation about evaluating an existing system and understanding what production readiness really involves.… Register for this event

LovableClaude

Sponsors: CGI

⌖ Frankfurt · byte5 GmbH

Am 13. Oktober 2026 baut das Claude-Meetup Frankfurt bei byte5 live eine neue Fähigkeit für Claude – mit Talks, Engineering-Praxis und Networking. Stop Prompting, Start Engineering Unter dem Motto „Skill Forge“ geht dieses deutschsprachige Meetup über einzelne Prompts hinaus: Im Mittelpunkt steht die Frage, wie sich für Claude eine neue Fähigkeit systematisch entwickeln lässt. Teilnehmende erleben den Entstehungsprozess live und erhalten praxisnahe Einblicke in Prompt Engineering, KI-Workflows und die Entwicklung agentischer Systeme. Das Treffen richtet sich an Entwickler:innen, AI Builder, technisch interessierte Produktmenschen und alle, die Claude nicht nur nutzen, sondern besser verstehen und gezielt erweitern möchten. Ein konkretes Erfahrungsniveau wird nicht vorausgesetzt; besonders relevant ist das Programm jedoch für Menschen, die sich bereits mit künstlicher Intelligenz, AI-Anwendungen, Webentwicklung oder Automatisierung beschäftigen. Vorläufiger Ablauf - 17:30 Uhr: Ankommen und Check-in - 17:45 Uhr: Vortrag 1 – Thema wird noch bekannt gegeben - 19:00 Uhr: Pizzapause und Networking - 19:30 Uhr: Vortrag 2 – Thema wird noch bekannt gegeben - 21:00 Uhr: Networking und Ende Die Kombination aus Live-Building, Vorträgen und persönlichem Austausch macht das Format besonders praktisch. Neben den fachlichen Sessions bleibt ausreichend Zeit, Erfahrungen zu vergleichen, eigene Claude-Projekte zu diskutieren und Kontakte in der lokalen AI-Community zu knüpfen. Getränke, Snacks und Pizza sind vor Ort vorgesehen. Die konkreten Vortragsthemen und Speaker stehen derzeit noch nicht fest. Wer selbst etwas präsentieren möchte, kann über das Call-for-Speakers-Formular einen Vorschlag einreichen. Themen aus der Community Bei früheren Treffen ging es unter anderem um Prompting, KI-Workflows, das Open-Source Agentic OS der Community sowie Omadia. Einen Eindruck von bisherigen Inhalten vermitteln die veröffentlichten Vortragsfolien von byte5.… Register for this event

Claude

◎ Online

Hosted by Packt Publishing Limited

Build, review and ship software with Claude Code in this live online bootcamp on 18 October 2026, from 2pm to 8:30pm UTC. Move beyond basic AI-assisted coding Claude Code Bootcamp Cohort 2 is an intensive, 6.5-hour hands-on workshop for developers who want structured, reliable and repeatable AI engineering workflows, not simply a longer list of prompts. Led by headliner Luca Berton and organised by Packt Publishing Limited, the session connects Claude Code’s individual capabilities into a practical system spanning codebase understanding, implementation, testing, review and preparation for production. Across 10 new advanced hands-on projects, you’ll explore how to: - Structure projects with CLAUDE.md and scoped rules - Create and apply reusable Claude Code Skills - Build specialised subagents for security and testing - Add hooks and guardrails that automate checks and block unsafe actions - Use Best-of-N workflows to compare AI-generated solutions - Test, review and verify AI-generated code - Assess whether work is ready to merge and ready for production - Develop more controlled and repeatable AI-assisted engineering processes Rather than treating these as isolated features, the bootcamp demonstrates how they work together across the software development lifecycle. The live, project-based format is designed to help attendees practise real workflows and leave with a more complete Claude Code toolkit. Who should attend? The programme is suitable for software developers, AI/ML engineers, platform and DevOps engineers, tech leads, developers exploring Claude Code, and people preparing for Claude certification. Basic programming ability and familiarity with Git are recommended, making this most relevant to technical attendees who already understand core development concepts but do not necessarily have extensive Claude Code experience. This is an independent Packt workshop rather than official Anthropic training. It is not endorsed by Anthropic, does not contain official examination questions, and completion does not guarantee success in an official certification exam. Refunds are available up to 10 days before the event. Get tickets for Claude Code Bootcamp Cohort 2

Claude

◎ Online

Hosted by AI Strategy Lab - Manchester

Build a Personal AI DNA file in this 45-minute online workshop and make ChatGPT, Claude, Copilot or Gemini produce writing that sounds like you. Hosted by Emery Layton for AI Strategy Lab – Manchester, this practical session takes place on Thursday 22 October 2026. It tackles a familiar problem for business owners and independent professionals: AI can produce polished copy, but the result often feels generic, anonymous and time-consuming to edit. Rather than treating this as a prompting failure, Emery frames it as an onboarding problem. During the workshop, you will create a Personal AI DNA file: a single page of plain text that gives an AI essential context about who you are, who you serve, how you communicate and what you would never say. What you’ll cover The session combines accessible AI strategy with a live, hands-on build. Topics include: - What AI is in plain language, and what it cannot do - Three current news developments that matter to business owners - Where AI tools are heading and what remains constant as they change - Why AI-generated writing often feels generic - The six components of a Personal AI DNA file, illustrated with examples - A four-minute method for capturing your writing voice - A live build in which Emery writes his file while you create yours - A side-by-side test using the same request with and without your DNA file You can use a free account for ChatGPT, Claude, Copilot or Gemini. No paid plan or specialist software is required; a laptop and a basic notepad are enough. Who it’s for This workshop is designed for business owners, coaches, consultants and freelancers who write frequently and want to reduce the time spent rewriting generic AI output. Complete beginners are welcome, while regular AI users should find particular value in the build and testing portion. Emery Layton is a licensed Professional Engineer with 25 years of experience in water and wastewater infrastructure. Working in a field where errors carry real liability shaped his approach to AI: accuracy, consistency and reliability take priority over novelty or showmanship.… Register for this event

ChatGPTClaudeGemini