Voxxed Days Zurich 2026
2026-03-24
Zurich, Switzerland
[Voxxed Days Zurich 2017, Voxxed Days Zurich 2019, Voxxed Days Zürich 2022, Voxxed Days Zurich 2023, Voxxed Days Zürich 2024, Voxxed Days Zürich 2025, Voxxed Days Zurich 2026]
who it's for
Java developers, platform engineers, machine learning engineers, security engineers, cloud architects, frontend developers, engineering managers
topics
Quarkus MCP applications; Spring AI multi-agent orchestration; federated learning with homomorphic encryption; always-on JVM profiling; Project Leyden and Spring Boot performance; Hotspot auto vectorization and Vector API; native SQL assertions; shielded VMs and workload identity; Model Context Protocol (MCP); privacy-first browser-based AI; JBang single-file Java tooling; database connection pool sizing
| title | author | length |
|---|---|---|
| Welcome to Voxxed Days Zurich 2026 | Federico Yankelevich, Patrick Baumgartner | 17m |
Ten-year anniversary starts with concrete changes: a fifth room added, five rooms to host more tracks; live transcription is running as an accessibility test… | ||
| The Promise of Trustworthy AI | César Soto Valero | 33m |
Claims that models improve without ever collecting raw data by combining Federated Learning with Homomorphic Encryption, demonstrating encrypted weight… | ||
| From Shadows to Spotlight - How Swiss Post Performs Reliable ML Deployment | Giovanni Degiorgi | 43m |
Swiss Post moves its Iris fraud service from research to production with a reproducible pipeline: data rooms on S3, SageMaker pipelines for training in a… | ||
| Create Intelligent MCP based applications with Quarkus | Dimitris Andreadis | 39m |
Quarkus ships a full Java MCP implementation that exposes annotated Java methods as MCP getCurrentTime tool endpoints and runs as local stdio, websockets or… | ||
| Agentic AI Patterns | Kevin Dubois, Mario Fusco | 47m |
Agentic orchestration uses LangChain4j with Quarkus to compose multiple specialized agents into concrete patterns: sequence, loop (max 5 iterations, exit ≥80%… | ||
| Secure By Design: Architecting cloud and AI workloads from the perspective of infosec | N. Godec | 37m |
Covers concrete controls to harden cloud and AI runtimes: enable secure boot and shielded VMs on node pools, make control planes private, use workload identity… | ||
| Data Integrity 2.0. The biggest advancement in data quality in 30 years! | Connor McDonald | 46m |
Demonstrates a working implementation of native SQL Assertions that enforce complex cross‑table rules declaratively: examples include “no employee salary above… | ||
| Finding the Spark: A Career Reset Button | Loïc Magnette | 42m |
A personal case study shows career apathy reverses by joining and contributing to a local community: attending meetups, submitting CFPs, designing a new… | ||
| Running Your Coding Agent Locally: Lessons from a Real-World Experiment | S. Maestri, A. Soldano | 45m |
Builds a local coding agent using Llama CPP and LM Studio with GGUF open weights, running quantized models (example Quen 3.5 122B at Q4_KXL) on consumer… | ||
| Social Engineering: Hacking Humans | Pawel Sucholbiak | 46m |
Lists major breaches tied to human tricks: Facebook $99M, Ubiquiti $39.1M, Google $23M, Toyota $37M, Puerto Rico $2.6M, Crelan $75.8M, plus a 2019 Tallahassee… | ||
| Coding a Multi-Agent Game Master with Spring AI | Alessandra Pasini, Arnaud Jean | 43m |
Live-coded a multi-agent Game Master that runs end-to-end: Spring AI + Java 25 + Gbang for agents, Amazon Bedrock (CloudIQ-4.5) as the LLM, and Spring Boot for… | ||
| Debugging Your Health: Fitness for Desk Warriors | Radek Kargul | 18m |
Sitting eight hours equals a season on Netflix and “almost two marathons” in duration, and sustained desk time causes hunching, weak upper-back muscles, tight… | ||
| Don’t Panic! A Guide to Keeping the Human Edge in a Universe of Instant Output | Janina Sutter | 20m |
Argues that AI accelerates output but erodes the tacit foundations that make work rewarding and creative, citing Dreyfus-style skill levels, Polanyi’s tacit… | ||
| Who to blame? The AI, The Programmer, or The Prompt? | Makan Sepehrifar | 43m |
Five live case studies show concrete bugs and fixes: a Copilot image proxy leak that exfiltrates secrets via camo URLs and leads to CVE-2025-53773, an… | ||
| Break the silo: How we built a decentralized control plane for service orchestration | E. Taborelli | 41m |
Julius Baer builds a decentralized control plane called Fleet Commander using Kubernetes + Crossplane (deployed on AKS), Flux for GitOps, provider-kubernetes… | ||
| Are We Ready For The Next Cyber Security Crisis Like Log4Shell? | Soroosh Khodami | 46m |
Live demo shows a reverse shell created by adding a crafted dependency to a Spring Boot project and by exploiting Maven/npm resolution — a combo of dependency… | ||
| The Hidden Security Hazards in Your Java Stack | Brian Vermeer | 48m |
Cloud‑native Java workflows produce concrete risks: an AI assistant (Copilot) outputs code that enables SQL injection via concatenated queries, Burp Suite… | ||
| AI without spaghetti: Clean architecture in the age of AI | Lutske de Leeuw, Maarten Vandeperre | 41m |
A recipe site called Leo becomes a chatbot and then a refactored platform that separates domain, use cases and infrastructure so models change without breaking… | ||
| Profiling Without Pain: Always-On Performance for the JVM | Scott Gerring | 40m |
A Quarkus feature-flag service with Postgres shows 1.48s traces and 10s GC-driven spikes until memory is increased; an N+1 Hibernate query is fixed with a JPQL… | ||
| Getting more out of Maven | Marit van Dijk, Andres Almiray | 42m |
Shows concrete Maven practices: prefer mvnw verify over install to avoid polluting ~/.m2, add a Maven wrapper (used 3.9.14 in demo) for reproducible builds… | ||
| The Ultimate Guide to Browser-Based AI: Privacy-First, Offline-Ready and Future-Proof | M. Salnikov | 46m |
Demonstrates running real ML inference entirely in the browser using WebNN as a low‑level bridge to CPU/GPU/NPU hardware (requires flags or origin trial), then… | ||
| Develop your own Browser extension | Łukasz Nowak | 20m |
Shows that a browser extension can be built like a frontend SPA and actually replaces repetitive dev tasks: a Chrome-only extension is created with Vite +… | ||
| Did You Not Get the Memo? Signed, Atomic Design | Faris Aziz | 21m |
Claim: combining Atomic Design with an understanding of React reconciliation cuts memo usage roughly in half and reduces needless re-renders. Evidence: a… | ||
| Ashes to Ashes, Specs to Specs: A Farewell to Old Testing Tools | Younes Jaaidi | 43m |
Shows how to replace Jasmine/Karma or Jest with Vitest on Angular apps using the Angular plugin and NX, demonstrating a progressive migration that keeps older… | ||
| Micro-Frontend & Me, a love story | Dan Neciu | 35m |
Argues that microfrontends actually solve one concrete pain: independent team deployments, proven by interviews with companies of 50–200 engineers and a… | ||
| JavaScript in 2026 | Christian Woerz | 43m |
Live coding shows concrete replacements for old patterns: use ES2023/24 array helpers like Array.prototype.toSorted and with instead of splice or in‑place… | ||
| Model Context Protocol (MCP): The future of user interactions? | Jonas Bandi | 45m |
Explains how the Model Context Protocol standardizes LLM access to services by registering tools with input/output schemas, shows a live webshop demo where an… | ||
| Hidden gems in Angular: hybrid rendering, new routing features, and beyond | Vojtech Mašek | 31m |
Shows how to cut boilerplate and actually ship faster: default-exported components let router imports drop the selector boilerplate, route redirects accept a… | ||
| Team and Software: Making an Impact as an Architect | Eberhard Wolff | 45m |
Argues that technical skill alone doesn’t move systems: architects must convert technical quality goals into decisions teams actually implement, so… | ||
| Database Connection Pool Sizing - Demystified! | Jasmin Fluri | 39m |
Claims that bigger pools always help are false: using Little's Law the example shows 50 tps at 100 ms needs ~5 connections, not hundreds, while a misconfigured… | ||
| Meeting Impossible SLAs: How we made our uptime 99.999% | Warren Parad | 46m |
Claims a 99.999% uptime SLA and demonstrates it holds during a recent US‑East‑1 DynamoDB/DNS outage by combining math, tooling, and procedures. Architecture… | ||
| Spring Boot Test - A look behind the curtain | Michael Vitz | 45m |
Shows why @SpringBootTest boots "95% of the app" (web server mocked) and when to prefer slices like @WebMvcTest; builds a reusable JUnit extension called demo… | ||
| Supercharge your JVM performance with Project Leyden and Spring Boot | Moritz Halbritter | 42m |
Project Leyden techniques plus Java 25 and Spring Boot cut a PetClinic startup from ~3.0s to ~0.8s using AOT cache (3.1x) and to ~0.075s with Spring AOT code… | ||
| From Fragile to Fearless: Writing Tests That Last | Wouter Bauweraerts | 46m |
Compares solitary (Mockito-heavy) and sociable unit tests with live code: a veterinary checkup price method is refactored while running 59 tests; sociable… | ||
| JBang, a Java file to rule them all? 💍 | Stéphane Philippart | 18m |
Demo shows JBang (installed via SDKMAN/brew/curl, version JBang 0.137) turning a ten‑line HelloWorld into a runnable script by building an executable jar on… | ||
| Building trust in a system through transparency: lessons from Swiss Post E-Voting system | Céline Camacho, Fabrizio Beretta Piccoli | 21m |
Swiss Post e‑voting publishes all source code and protocol docs, enforces strict code conventions, two‑approval PRs and cryptographic reviews, and runs Sonar… | ||
| Quarkus Unleashed | Roberto Cortez | 50m |
Quarkus shifts expensive runtime work into compilation so apps start fast and use less memory, achieved by build steps and build items in extension deployment… | ||
| Reactive ❤️ Loom: A Forbidden Love Story | Francesco Nigro | 46m |
Quarkus and Netty experiments show a custom Loom scheduler that runs virtual threads on the same platform threads as Netty event loops, implemented in ~400… | ||
| Reversible Decisions for Better Software Design | David Tanzer | 44m |
Claim: making decisions reversible keeps long-lived projects fast to change by containing coupling and preserving options. Practical moves include creating a… | ||
| Hotspot JVM Auto Vectorization and Vector API | Emanuel Peter | 44m |
C2’s SuperWord auto‑vectorizer packs unrolled loops into SIMD and on the normal‑map demo a loop goes from ~36 FPS to ~130 FPS; a Vector API N‑body demo reaches… | ||
| Aspects of Continuous Deployment – A Multi-Dimensional View | Nadine Broghammer, Marc Sallin | 42m |
Swiss Post’s parcel sorting decision system runs in AWS Frankfurt on AKS with Kubernetes, ~20 services (mostly containers), .NET, Kafka (≈70 topics) and MQTT… | ||
| nothing matches | ||