AI in azienda: AI sotto controllo
Closing evening of the Ticino series on conscious AI use in business, this one on data, responsibility, rules and governance — content, a hands-on workshop and networking, capped at thirty participants.
25 upcoming workshops across Switzerland, checked daily.
Closing evening of the Ticino series on conscious AI use in business, this one on data, responsibility, rules and governance — content, a hands-on workshop and networking, capped at thirty participants.
Valais SMEs see how AI agents can take over routine admin — client requests, bookings, contracts and follow-up — with a live tool demo and a local case study from the Sion coworking space.
Lunchtime session with the Wyss Center AI Hub, SOPHiA GENETICS, the Swiss Data Science Center and BrainQuant on the data, infrastructure, validation and regulatory groundwork behind AI in research partnerships, precision medicine and digital biomarkers.
One-day SDSC training on designing, deploying and operating ML systems at scale — version control, CI/CD, experiment tracking, monitoring, drift detection and automated retraining. Registration closes 24 September.
An interactive executive lab run with Zeix on redesigning work around AI: deciding what the machine should actually do, and marking out where human judgement, responsibility and control have to stay.
A hundred students and early-career researchers spend a weekend on real pharmaceutical manufacturing problems, combining manufacturing data, mechanistic modelling and machine learning with industry and academic mentors before pitching to a jury.
A 90-minute hands-on build where everyone ships a working web app with AI features using Lovable’s built-in models — no API keys or separate accounts — with mentors on hand, then demos and drinks. Bring a laptop and an idea.
A full German-language day with Michael Kyburz on getting AI to take over recurring work: three base skills, how to build your own, and wiring assistants into calendar, mail and project tools so multi-step workflows run with clear checkpoints for human review. Fifteen places, and a paid ChatGPT or Claude subscription is required.
A German-language follow-on course for public-sector staff who already use generative AI daily: controlling output more precisely, building reusable prompt templates, and setting up your own CustomGPT or Copilot agent against your own use cases.
The on-site leg of the national Apertus hackathon: build on the open Apertus 8B model, strengthen its multilingual abilities and ship applications for private- and public-sector problems, with everything released as Open Commons.
Executive workshop for CEOs, CFOs, CDOs and CTOs on turning data and AI into measurable business value, with case studies from EKZ, Merck and Richemont. Limited seats.
Two-hour on-site workshop from the EPFL Library comparing four AI-powered research assistants, how they work and where they differ from traditional information retrieval. Open to the public, registration required.
A 90-minute German-language session for SME decision-makers on steering generative AI use rather than letting it spread: setting a strategic frame, drafting an internal AI policy, weighing data-protection, liability and wrong-output risks, and planning a safe rollout under Swiss data protection law and the EU AI Act. Fifteen places.
A two-day German-language course leading to the A4Q AI Foundation certificate, covering how AI and machine learning systems work, where they fit in a business, and the risks and limits that come with them.
The Basel leg of the national Apertus hackathon series, building applications and multilingual improvements on Switzerland’s open Apertus model at FHNW Campus Dreispitz, with everything produced released for reuse.
An in-person hackathon where mixed teams of students, developers, startups, public-sector innovators and corporate experts build practical AI solutions to challenges set by partner organisations, open to all skill levels.
Festive team hackathon where the Zurich Databricks User Group turns a sabotage mystery into hands-on practice with Genie, Lakebase and Databricks Apps.
A one-day hands-on event where participants try to reproduce one another’s analyses, papers and tools, then compare notes on what makes computational work hold up; the SwissRN annual meeting follows the next day at the same venue.
One-day German-language course for senior engineers, tech leads and CTOs who already code with agentic tools: build your own eval suite, an MCP tool wired to your build and test commands, a CI hook that blocks bad AI patches, and a cost-per-feature model comparison — working on your own repository. Limited to 16 participants.
A three-day course in Wengen on using generative AI responsibly in research, pairing lectures with computer practicals on analysis, literature searching, scientific writing, study websites and agentic AI — and on where error, bias and reproducibility break down.
A five-day English-language executive course for leaders who need to challenge assumptions on AI projects without becoming data scientists — data science foundations, MLOps, AI governance and ethics, generative AI and a first agent, with an ETH Zurich certificate. Registration closes 15 January 2027.
Two-day German-language course for developers, architects, data scientists and ML engineers on building agentic LLM systems that survive production: function calling, hybrid-search RAG, LangGraph topologies, security hardening and systematic evaluation. Places are limited.
An afternoon course run with Quantum Basel, led by its CTO Jan Mikolon, bridging current AI practice and quantum technologies — from the foundations through to what it takes to scale.
A two-day hackathon in Geneva, run as an official side event of the Applied Machine Learning Days, where mixed teams of Pictet employees, students and external builders turn real financial-services challenges into working agentic-AI prototypes; applications close on 7 January 2027.
One-day German-language course for senior engineers, tech leads and CTOs who already code with agentic tools: build your own eval suite, an MCP tool wired to your build and test commands, a CI hook that blocks bad AI patches, and a cost-per-feature model comparison — working on your own repository. Limited to 16 participants.
No events match those filters.