Brainberg

AI & ML Research Events in Europe

Europe's AI and ML research scene is a long tail of university labs, community research groups, and practitioner-run meetups that together punch well above their weight globally. This page covers the research end of the AI/ML event calendar: deep-learning meetups, computer vision and NLP groups, paper-reading clubs, and research-oriented conferences. It's aimed at ML engineers, applied researchers, and PhD students who care about how the models work, rather than how to integrate them into a product (that's the Applied AI category).

Anchor events include the MLcon series (Berlin, Munich, Amsterdam, London), PyData conferences, and regional deep-learning meetups like the Vienna Deep Learning Meetup. Quantum AI and quantum-computing events sit here too, since the European quantum community is research-heavy and overlaps meaningfully with the ML research crowd. Topics cover training and serving frameworks, fine-tuning technique, evaluation, quantization, model architectures, and the infrastructure that makes experimentation tractable.

Brainberg aggregates these into a single chronological European view. For the deluge of "how to ship a feature with an LLM" events, see the Applied AI category instead.

Upcoming events

AI/ML Research & EngineeringMeetupFree

75th Deep Learning Meetup: AI Manipulation / Scaling LLM Infrastructure

Vienna, 🇦🇹 Austria

Hi Deep Learners,

We are happy to announce our first Vienna Deep Learning Meetup after the summer break: On September 23
we'll be hosted by 42 Vienna (in Heiligenstadt) and will feature two exciting topics:

  • Misaligned AI agents (on the example of recent Open AI incidents)
  • Scaling AI Infrastructure on premise

***
Agenda:

  • 18:15 Arrival
  • 18:30 Welcome by the meetup organizers
  • Introduction by the host: 42 Vienna
  • 18:45 Talk 1: Misaligned AI agents are here: how to evaluate frontier models that collude and know they are being tested by Jason Hoelscher-Obermaier (Director of Research at Apart Research)
  • 19:30 Announcements
  • Networking Break
  • 20:00 Talk 2: AI Inference Engines and Instances: Strategies for Scaling LLM Infrastructure by Jonas Aaron Vander (CTO at Xinity)
  • 20:30 Networking
  • ~22:00 Wrap up & End

***

Talk Details:
--------------
Talk 1: Misaligned AI agents are here: how to evaluate frontier models that collude and know they are being tested
In May 2026, OpenAI agents working on a timed web-lookup task found DSEwiki, a German-language developer wiki run from Graz since 2001. These agents were sandboxed to only have read-access to the internet but were able to edit the wiki through GET requests. Over four weeks they left about 17,000 edits, shared answers for their tasks and collaborated on ways to score more highly. Over the same weeks and into July, roughly 1,200 sandboxed OpenAI agents working on ExploitGym evaluation tasks colluded via a covert message board and about 700 took part in breaking into Hugging Face.
In this talk, I will cover what happened and why, and offer takeaways if you build or evaluate agents. In short: misaligned agents are real; they understand they are being evaluated and attempt to game the evaluation; and even sandboxed agents should not be assumed safe.
I will close on what could actually be done about this, drawing on the manipulation evaluations we built for AI Act enforcement by the EU AI Office over the past year: what we need for frontier AI evaluations to be reliable indicators of risk, and how far AI Act enforcement can help when incidents happen with pre-deployment models.

About the speaker:
Jason Hoelscher-Obermaier is Director of Research at Apart Research, based in Vienna, where he has led the organisation's work on AI evaluations, including manipulation-risk evaluations for the EU AI Office. He holds a PhD in physics from the University of Vienna, worked on dangerous-capability evaluations at ARC Evals (now METR), and was an ML engineer at two European AI startups.

Talk 2: AI Inference Engines and Instances: Strategies for Scaling LLM Infrastructure

LLM inference is more than just deploying a model. With Ollama, vLLM, SGLang, there are a variety of specialized engines, each with its own strengths in latency, throughput, ease of use, and configuration. In this talk, we take a practical look at modern LLM infrastructure: Which engine to use when? How do we scale beyond single nodes? And why is context-aware orchestration the key to efficiency when connecting different engines in a cluster?
By the end, you’ll have a roadmap for choosing the right engine for your use case and managing it in a scaled production environment.

Section topics (tentative):

  • Modern LLM infrastructure is multi engine
  • Ollama excels at local and edge efficiency
  • vLLM dominates high-throughput production serving tasks
  • SGLang is ideal for structured outputs
  • TGI suits HuggingFace ecosystem integration
  • Generic routers suffer from cache blindness
  • Context-aware orchestration is the future

About the Speaker
Jonas Aaron Vander is CTO and co-founder of Xinity, a Vienna-based sovereign AI infrastructure company, and the architect behind Xinity Runtime, an open-source, OpenAI-compatible inference platform that lets regulated European enterprises run LLMs entirely on their own hardware, currently serving production workloads like Mediengruppe Wiener Zeitung. His background is in AI solution architecture and MLOps.

We are looking forward to welcoming you at this meetup!
Your VDLM organizer team

Wed 23 Sept · 16:3050–200
AI/ML Research & EngineeringMeetupFreeOnline

Prompting Lab: "Prompting mit Knowledge Graph: Hochgenaue, erklärbare Antworten"

📅 Termin: Freitag, 25.09.2026, 13.15-14.00
ℹ️ Thema: „Prompting mit Knowledge Graph: Wie Du hochgenaue, erklärbare KI-Antworten bekommst“ mit Dr. Sebastian Goeser
Wenn die KI nicht weiß, woher sie ihr Wissen nimmt, rät sie. Sebastian Goeser von der Textverstehen GmbH arbeitet an genau dieser Stelle: mit Knowledge Graphen als sauberer Wissensbasis für Prompting. Er zeigt, wie so ein Graph entsteht, wie man darin zur passenden Information navigiert und welche Rolle das Sprachmodell am Ende überhaupt noch spielt. Nebenbei wird klar, warum es dafür oft gar keinen großen Anbieter braucht, sondern ein kleines lokales Modell reicht. Nach einer kurzen Einordnung geht es direkt in die Live-Demo.

Was passiert bei unseren Sessions?
Unsere Veranstaltungen bieten dir kompakte Einblicke von erfahrenen Experten – darunter auch Profis renommierter Unternehmen wie dem Handelsblatt. Statt langer Theorie erwarten dich praktische, sofort anwendbare Tipps und Beispiele, die dich wirklich weiterbringen. Bei uns wird es nie trocken oder langatmig – ob Einsteiger oder Fortgeschrittener, hier findet jeder neue Inspiration.

Darauf kannst du dich freuen:
• Neue Prompting-Tricks und Best Practices: Konkrete Ansätze, die du sofort in deinen Alltag integrieren kannst.
• Experimente in einem geschützten Raum: Probiere dich aus – ohne Druck oder Scheu.
• Konstruktive Prompt-Reviews: Erhalte Unterstützung bei Fragen oder Herausforderungen.
• Austausch und Networking: Vernetze dich mit Gleichgesinnten, die deine Begeisterung für KI teilen.
• Aktuelle Trends und Anwendungsbeispiele: Bleib up-to-date und gestalte den Fortschritt aktiv mit.

❓Warum solltest du dabei sein?
Werde Teil unserer wachsenden Community und gestalte die Zukunft des Promptings mit. Bei uns zählt: Wo du bist, ist der richtige Startpunkt. Mach mit und werde Teil dieser spannenden Entwicklung!

Du bist schon ein Prompting-Expert:in?
Perfekt! Teile deine Erfahrung, präsentiere deine besten Ideen und Tricks, und inspiriere andere. Wir freuen uns auf deinen Input und darauf, von dir zu lernen

Probleme beim Eventbeitritt?
Schreibt uns eine Mail an: events@safari-consulting.de

Fri 25 Sept ¡ 11:15< 50
AI/ML Research & EngineeringMeetupFree

Sept 25 - Berlin Physical AI, ML, and Computer Vision Meetup

Berlin, 🇩🇪 Germany

Join our in-person meetup to hear talks from experts on cutting-edge topics across AI, ML, and computer vision.

Date, Time and Location

Sep 25, 2026
5:30 PM - 8:30 PM CEST
w3.hub, Möckernstraße 120, 10963 Berlin, Germany

One Pipeline, Any Device: Swap the Reader, Not the Pipeline

Every new robot, camera, or sensor vendor in physical AI usually means rebuilding the data pipeline from scratch, even though the features you actually need, a calibrated frame, a fused object distance, stay the same. This talk demos mloda, an open-source Python framework where a feature pipeline is written once, and only the reader plugin, the small piece that knows how to read one specific device's raw format, ever changes.

Live, using two small synthetic datasets shaped like two different devices' raw output (no hardware involved), I run the same pipeline against both, swapping only the reader plugin, and show the calibration, fusion, and feature extraction steps running unmodified on either. I also show OpenTelemetry lineage tracing, so when a feature looks wrong, you can trace it straight back to the raw reading that produced it, regardless of which device it came from.

The point: stop rebuilding your data pipeline for every new device, reuse it.

About the Speaker

Tom Kaltofen is a Berlin-based data and AI engineer and the creator of mloda, an open-source (Apache-2.0) Python framework for declarative, plugin-based data access in AI workflows.

Toward the best ROI: choosing algorithms for AI-powered workstations

Robots are becoming smarter and more affordable—but which automation projects actually deliver a return on investment? Drawing on RemBrain’s real-world experience across delivery, retail, construction, and manufacturing, this presentation reveals why many promising robotics concepts fail to become viable products.

It introduces a practical, skill-based approach to flexible automation and shows how compact AI-powered workstations can achieve payback in as little as 6–12 months. The talk also explores where Vision-Language-Action models create genuine value—and where simpler, proven technologies remain more effective.

Honest, numbers-driven discussion with an algorithm focus.

About the Speaker

Anton Maltsev Started working with Computer Vision in 2010. From 2017 to 2022, I was Head of ML at Cherry Labs, which was acquired by Artisight. Now CSO at Rembrain.

Building Real-World Computer Vision Systems

This talk will explore practical workflows for building, evaluating, and improving modern computer vision systems. We'll dive into real-world approaches to dataset curation, model analysis, multimodal AI workflows, and production-ready vision pipelines using open-source technologies.
The session is designed for engineers, researchers, and AI practitioners looking to better understand how teams are developing and scaling computer vision applications today. Expect practical demos, technical insights, and discussions around the evolving AI tooling ecosystem.

About the Speaker

Dan Gural leads technical partnerships at Voxel51, where he's building the Physical AI Workbench, a platform that connects real-world sensor data with realistic simulation to help engineers better understand, validate, and improve their perception systems.

Fri 25 Sept · 15:3050–200
AI/ML Research & EngineeringMeetupFreeOnline

Introduction to Machine Learning in Epidemiology with R

How can machine learning help us work with epidemiological data? And how do we know whether a predictive model is actually useful?

Join R-Ladies Rome for a practical, two-hour introduction to machine learning in epidemiology using R.

In this workshop, Federica Gazzelloni will introduce the main ideas behind supervised machine learning through an applied epidemiological example. Rather than focusing on a long list of algorithms, we will follow the complete machine-learning workflow: from defining an epidemiological question to training, evaluating and interpreting predictive models.

We will explore how different models approach the same prediction problem, starting with logistic regression as a baseline and moving to decision trees and random forests.

Using R, we will look at how to define a classification task, train models, generate predictions and evaluate their performance on unseen data.
Particular attention will be given to model evaluation and interpretation.

Throughout the workshop, we will also consider an important distinction for epidemiological research:
Prediction is not the same as inference, and predictive importance does not imply causation.

The workshop is inspired by the recent Machine Learning in Epidemiology study by Wright et al. (2026) and connects with Federica's book, Health Metrics and the Spread of Infectious Diseases: Machine Learning Applications and Spatial Modelling Analysis with R (CRC Press, 2025), where she introduces machine-learning applications in health and infectious-disease research using the mlr framework.
During the workshop, we will use the modern mlr3 ecosystem and discuss how machine-learning workflows in R have evolved from mlr to mlr3.

What we will cover

  • What machine learning means in an epidemiological context
  • From an epidemiological question to a prediction task
  • Preparing data for machine learning
  • Logistic regression as a baseline model
  • Decision trees and random forests
  • Training and evaluating models with mlr3
  • Cross-validation and performance on unseen data
  • Sensitivity, specificity, confusion matrices and ROC/AUC
  • Variable importance and model interpretation
  • Prediction versus explanation and causation
  • Limitations, bias and data quality in epidemiological machine learning

Who is this workshop for?

The workshop is designed for R users interested in epidemiology, public health, health data or machine learning. Basic familiarity with R and data analysis is useful, but no previous machine-learning experience is required.
The session will combine explanation, live R coding and discussion, with an emphasis on practical and reproducible analysis.

About the instructor
Federica Gazzelloni is an actuary, statistician, data scientist, author and instructor, and the founder and organiser of R-Ladies Rome. Her work spans health metrics, statistical modelling, machine learning, reproducible research and R.
She is the author of Health Metrics and the Spread of Infectious Diseases: Machine Learning Applications and Spatial Modelling Analysis with R, published by CRC Press in 2025.

R-Ladies
R-Ladies is a worldwide organisation promoting gender diversity in the R community. R-Ladies Rome provides a welcoming space to learn, share knowledge and connect with people interested in R, data science, statistics and reproducible research.
Everyone is welcome to attend, regardless of gender identity or level of experience.

Mon 28 Sept ¡ 16:00< 50
AI/ML Research & EngineeringMeetupFree

31st Belgium NLP Meetup

Ghent, 🇧🇪 Belgium

It's back to school week, and that means meetups, too, spring back to life. With agents that escape their sandboxes, a blistering pace of model releases, and a couple of landmark IPOs looming, NLP and AI enthusiasts certainly have enough to talk about.

The next Belgium NLP Meetup will take place on Thursday October 8th in the offices of Trendtracker in Ghent. We'll open the doors at 7pm for pizza, have three interesting talks between 7.30pm and 9pm, and after that there's more time for networking and drinks.

Below are the three talks for the evening:

Relex: Relationship Extraction at Scale for Trendtracker's Knowledge Graph
François Remy (Trendtracker)
What can you do when you need to extract relationships at such a scale that LLMs stop being economical? In this talk, François will explain how Trendtracker trained its own multilingual relationship extraction model to power its knowledge graph, which jointly identifies entities and links them by relationship type. He'll discuss why they chose a bi-affine architecture, and how scale, cost, and reproducibility shaped the design. He'll also show how the extracted relationships can be leveraged to help with entity resolution.

Open Source the LLM Engineer, not just the LLM
Sofie Van Landeghem (OxyKodit)
This talk starts rather innocently introducing nanochat: a fully open,
human-readable implementation of a simple LLM with GPT2 capabilities. Sofie then uses nanochat as an example to demonstrate how a bigger LLM can turn into an LLM engineer in an "autoresearch" loop. Finally, she'll show an early prototype of such an open-source LLM engineer, built by a human LLM engineer (Sofie), let loose on a GPU cluster to perform genuinely open AI research.

Beyond CLAUDE.md & .cursorrules: Architecting GitOps and Context Governance for AI Coding Agents
Younes Baghor (SPRAWL.software)
Autonomous coding assistants like Cursor, Claude Code, Codex and GitHub Copilot have dramatically accelerated developer velocity. But as engineering teams adopt these tools across complex codebases, they are hitting a frustrating architectural wall: "Agent Sprawl." In this talk, Younes Baghor, Head of Agentic AI, Systems Architect, and the Founder of SPRAWL.Software, explores how to apply traditional software engineering discipline to AI coding workflows. He’ll do a technical deep-dive into replacing fragmented, ad-hoc prompt files with a unified GitOps architecture for AI agents.

Thu 8 Oct ¡ 17:00< 50
AI/ML Research & EngineeringConferenceFree

ICMLT 2027 in Stockholm - Machine Learning Technologies

Stockholm, 🇸🇪 Sweden

Get ready to dive into the latest machine learning tech and connect with fellow geeks at ICMLT 2027!

Abbreviation: ICMLT 2027

Full Name: 2027 12th International Conference on Machine Learning Technologies (ICMLT 2027)

Location: Stockholm, Sweden

Conference Date: May 21-23, 2027

Official Website: www.icmlt.org

Publication

All registered and presented papers will be published into ICMLT Conference Proceedings, which will be submitted for Ei Compendex, Scopus, and other databases.

ICMLT 2025 丨IEEE Xplore丨 ISBN: 979-8-3315-3672-5丨Ei Compendex and Scopus
ICMLT 2024 丨ACM Digital Library丨 ISBN: 979-8-4007-1637-9丨Ei Compendex and Scopus
ICMLT 2023 丨ACM Digital Library丨 ISBN: 978-1-4503-9832-9丨Ei Compendex and Scopus
ICMLT 2022 丨ACM Digital Library丨 ISBN: 978-1-4503-9574-8 丨Ei Compendex and Scopus
ICMLT 2021 丨ACM Digital Library丨 ISBN: 978-1-4503-8940-2 丨Ei Compendex and Scopus
ICMLT 2020 丨ACM Digital Library丨 ISBN: 978-1-4503-7764-5 丨Ei Compendex and Scopus
ICMLT 2019 丨 ACM Digital Library丨ISBN: 978-1-4503-6323-5 丨Ei Compendex and Scopus
ICMLT 2018 丨 ACM Digital Library丨ISBN: 978-1-4503-6432-4 丨 Ei Compendex and Scopus

Call for Paper

Adaptive systems

Business intelligence

Biometrics

Data and web mining

Neural net and support vector machine

Pattern Recognition

Hybrid and nonlinear system

Intelligent and knowledge based system

Fuzzy set theory, fuzzy control and system

More topics via: http://www.icmlt.org/cfp.html

Submission Methods

Full Paper (publication and oral presentation)

Abstract (oral presentation only)

Electronic Submission System (.pdf)

http://confsys.iconf.org/submission/icmlt2027 or email: icmlt_conf@163.com

Contacts

Ms. Stacy Lee

E-mail: icmlt_conf@163.com

Tel: +86-13096333337

Fri 21 May 12:00 – Sun 23 May 20:00