GPU, CUDA, and PyTorch Performance Optimizations

⌖ Berlin · Park Inn by Radisson Berlin Alexanderplatz Hotel

Hosted by Global AI Berlin

KanDDDinsky ’26 brings Domain-Driven Design and AI-assisted software development to Berlin from 14–16 October 2026 for talks, hands-on sessions and Open Space. Now in its ninth edition, KanDDDinsky is a community-led conference for people who build business software and want to improve how they design, develop and evolve it. Around 300 attendees will gather at the Park Inn by Radisson Berlin Alexanderplatz Hotel in Berlin to learn from world-renowned experts, hear fresh community voices and exchange practical experience with peers. This is deliberately not a vendor conference. The emphasis is on collaborative learning, useful ideas and real-world software development rather than product pitches. The 2026 programme will place particular focus on AI-assisted software development in the context of Domain-Driven Design, exploring this fast-moving area through the conference’s established community format. Format and schedule KanDDDinsky ’26 combines several ways to participate: - Conference talks and hands-on sessions: 14–15 October - Open Space day: 16 October - Pre-conference workshops: 12–13 October The main event listing runs from 08:00 on Wednesday 14 October until 16:00 CEST on Friday 16 October. Talks offer access to expert and community perspectives, while hands-on sessions and workshops create opportunities to engage more directly with the material. The Open Space day gives participants a more collaborative setting in which the community can shape discussions around the subjects that matter most to them. Who should attend? The event is designed for software practitioners interested in Domain-Driven Design, artificial intelligence and the challenges of building effective business software. It will be particularly relevant to developers, architects and other technical professionals who want to examine AI-assisted development through the lens of software design and business context. The mix of expert-led content, practical sessions and participant-driven discussion should suit both established DDD practitioners and people looking to deepen their understanding through contact with the wider community.… Register for this event

Sponsors: Microsoft · Neo4j

⌖ Toronto · MaRS Waterfront, Lake Lounge

Impact AI Toronto brings enterprise AI leaders to Toronto on 14 October 2026 for practical strategies on governance, adoption and AI agents. Taking place from 10:00 AM to 3:00 PM at MaRS Waterfront’s Lake Lounge, this one-day conference explores how established organisations can move beyond AI experimentation and create lasting operational change. Across expert talks, candid panels and networking sessions, attendees will hear how leaders approach trusted data, workforce concerns, business-model transformation and institutional knowledge capture. What the programme covers The third Toronto edition opens with Hugues Foltz, Executive VP at Vooban, before moving into a practical look at AI governance. Gabriella Chidiebere, Head of Data Management at TD Bank, will explain what data models should, and should not, consume, how organisations can validate their inputs and the essential practices behind trustworthy AI. The human impact of adoption then takes centre stage. David Cachia, CIO at the Municipality of Clarington, will discuss leading change when employees fear job losses and layoffs are not an option. His session focuses on presenting AI as a way to augment skills, addressing scepticism and building adoption through direct conversations. Afternoon sessions examine AI at an organisational level: - When AI Rewrites the Business Model: Kevin Dreyer, CIO/CISO at Maple Reinders, joins a panel on defining the right problem, avoiding overhyped solutions and making AI stick inside legacy enterprises. The discussion will consider why data foundations matter more than models, and why redesigning the business can be more valuable than reducing headcount. - Capturing Expertise to Train AI Agents: Eric Whaley, CIO at Wolseley, and Carl Chouinard, CPO at Vooban, explore how to preserve knowledge held by experienced employees before it leaves the organisation. Expect discussion of what to record, what to automate and what should remain human. Who should attend?… Register for this event

◎ Online

Hosted by HUMAN + AI Interaction Science Lab ldn

Learn how to combine deterministic logic, machine learning and generative AI to create more reliable applications in this practical online session on 14 October 2026. Hosted by HUMAN + AI Interaction Science Lab ldn, the talk examines why not every part of an AI system should be left to generation. Using the ecosystem.Ai Workbench, the session will demonstrate how predictable rules can work alongside flexible generative capabilities to produce responsive interactive systems. What the session covers - The respective roles of deterministic and generative AI - Combining rules, machine learning and generative capabilities - Configuring different approaches within the ecosystem.Ai Workbench - Building AI applications that balance flexibility with predictable behaviour The event is relevant to people interested in artificial intelligence, machine learning, data science, predictive analytics and human-computer interaction. It takes place online on Wednesday 14 October 2026 at 3:00 PM; the timezone is not specified on the published page. The joining link will be visible to attendees. No detailed agenda or named speaker lineup has been published. Participants are asked to complete the participant registration. View the event and register on Meetup.

⌖ Amsterdam · Xebia Netherlands BV

Hosted by Xebia Software Development

Advanced AI Amsterdam is a free, in-person evening of technical talks on explainable workflows, agent memory and data-grounded AI agents on 14 October 2026. Hosted by Xebia Software Development at Xebia Netherlands BV in Amsterdam, this meetup brings together people who want to deepen their understanding of generative AI, large language models and agentic systems. Expect practical ideas from industry experts, opportunities to meet fellow innovators and engineers, and time to continue the discussion over complimentary pizza and drinks. Three perspectives on building better AI agents The evening features three focused sessions: - Explainable Agentic Workflows, Arman Kavian: Agents may narrate their reasoning without accurately reporting what drove an outcome. This talk explores preserving reasoning as structure so that an agent’s account can be checked against the process and evidence behind its result. - Can Agents Dream?, Florian Buetow: Most agent memory systems retrieve stored information on demand but do not replay and reorganise past actions. This session introduces “dreaming” as a memory architecture for improving long-running agents, outlining the main approaches and comparing them with retrieval-based memory systems. - Grounding AI Agents in Your Data Model, Ricardo Granados Lopez: See a working pipeline that grounds an AI agent’s claims about a data model in real evidence. The live walkthrough examines when AI-generated interpretations of SQL can be trusted automatically, and where they still require scrutiny. Schedule and format - 17:30: Walk-in, drinks and pizza - 18:00: Explainable Agentic Workflows - 18:40: Can Agents Dream? - 19:20: Grounding AI Agents in Your Data Model - 20:00: Drinks and networking - 21:00: Doors close This is a talk-led meetup with a live technical demonstration and informal networking. It is particularly relevant to AI engineers, software and data practitioners, agent builders, and technically curious attendees exploring prompt engineering, LLMs or generative AI. No formal prerequisite is listed, although the subjects are designed for people ready to engage with the architecture, reliability and evaluation of production-minded AI systems. View the event and registration details

⌖ Madrid · Madrid, Spain

Hosted by PyTorch

PyTorch Meetups Madrid launches its local AI and machine learning community with a free, in-person gathering in Madrid. Taking place on 14 October 2026, from 6:00pm to 8:30pm local time (CEST), the meetup creates a space for attendees to discuss PyTorch, modern deep learning and generative AI while connecting with others working across the field. The community welcomes a broad range of experience levels, including: - Engineers deploying machine learning models in production - Researchers working at the forefront of AI - Founders and builders starting new projects - Beginners taking their first steps in machine learning The emphasis is on sharing real-world experience, resolving technical questions and building collaborations across Madrid’s AI ecosystem. This event marks the official launch of the local PyTorch meeting point, with future community updates available through the PyTorch Madrid website. The event is free and will be held at Paseo de la Castellana, 259C, Madrid 28046, Spain. No detailed agenda or speaker line-up has been published yet. Register for the PyTorch Madrid meetup

⌖ Oxford · Old Fire Station

Hosted by TNG Technology Consulting UK

AI in the Wild is a three-hour evening of talks on value-generating, real-world AI applications at the Old Fire Station in Oxford on 14 October 2026. Hosted by TNG Technology Consulting UK, the event moves beyond speculation and hype to examine how AI is being implemented in practice. Through live case studies and insights drawn from different industries and business domains, speakers will explore intelligent systems that deliver tangible value and address complex organisational challenges. What to expect The programme centres on focused talks and practical discussion rather than abstract predictions about AI. Attendees can expect: - Deep dives into live, real-world AI case studies - Examples of value-generating AI applications across diverse business domains - Insights from practitioners with experience of AI implementation - Discussion of what actually works when intelligent systems meet real operational needs - Networking with people interested in artificial intelligence, business and technology The event is particularly relevant to IT and technology professionals, business audiences, and anyone interested in the realities of deploying AI. It should also appeal to practitioners working with production AI or using intelligent systems to solve demanding business problems. The emphasis is firmly on lessons from the field: what has been built, where it creates value and what others can learn from those implementations. Format and venue This is an in-person evening gathering combining a series of talks with opportunities for conversation and networking. It runs from 7:00pm to 10:00pm BST on Wednesday 14 October 2026 at the Old Fire Station, 40 George Street, Oxford OX1 2AQ. Call for speakers TNG Technology Consulting UK is also welcoming external contributors. If you have deployed AI in production or used intelligent systems to tackle a complex business challenge, the organiser would like to hear your proposed talk. Potential speakers can contact events-uk@tngtech.com with a topic idea. Come prepared to compare practical experiences, hear evidence from multiple industries and discuss where AI is already producing results in the wild. Register for AI in the Wild