Cloud and AI architecture
I help teams structure Azure and AI solutions so they are understandable, secure, and realistic to operate. The interesting work usually starts after the demo.
Senior Cloud Solution Architect · Azure · AI · data
I build and explain cloud, data, and AI systems.
I work on cloud, data, and AI systems, mostly where ideas need to become something teams can actually run.
Cloud architecture that can be explained
AI use cases that survive reality
Projects I keep learning from
How I usually help
Turning ideas into systems people can understand and use
I like working on the messy middle: when an idea sounds promising, but the team still has to figure out the data, architecture, security, cost, user workflow, and actual delivery path.
Data and machine learning
I have worked across many ML and data science use cases. That breadth helps me connect architecture decisions with the actual data, models, and user workflows behind them.
Products and experiments
I like building small, concrete products and experiments. They are the fastest way to find out which assumptions are solid and which ones only worked on slides.
Projects
Projects I worked on over the past years
Agents
Cronosaurus
Stopped
An experiment around scheduled agent runs, heartbeats, monitoring, and more complex autonomous tasks.
agentsautomationorchestrationscheduled-runs
AzureAgentic workflowsScheduled agentsHeartbeat monitoringTypeScriptPython
Content / Education
Relataly
Active
A blog I started while working as a machine learning consultant, with practical articles about ML, data science, and analytics use cases.
writingeducationtutorialsmachine-learning
Content systemsAI/MLDataSEO
AI Products
Neubaufee
Stopped
A real-estate data and AI pipeline for exploring new-building projects, subprojects, and project imagery.
productreal-estatescrapingmultimodalimages
AzureReactScraping pipelinesMultimodal modelsImage classification
Experiments
AI Use Case Hub
Exploring
A growing collection of practical AI use cases, demos, and implementation patterns.
use-casesaicatalogexperiments
Azure AIKnowledge modelingContent systems
Now
Current work
Current workUpdated May 24, 2026
More recently I have been exploring algorithmic portfolio management: how to structure signals, evaluate strategies, and keep the system explainable instead of turning it into a black box.
Based on 3 exploration notes.
Current workUpdated May 24, 2026
I have also been looking at agentic legal use cases on Swiss legal data, especially how retrieval, reasoning, and citations could support legal research workflows.
AI Use Case Hub
Based on 2 exploration notes.
Writing
Notes, articles, and project updates
project-updateApr 19, 2026
Launching a project-led home for current work
I rebuilt this site around projects I worked on over the past years, plus current experiments and writing.
Cronosaurus
project-updatewebsiteagents
writingApr 12, 2026
Agentic systems are moving from concept to operating surface
My writing and experiments increasingly focus on the practical side of agentic systems: state, boundaries, observability, and usefulness.
Relataly
writingagentsarchitecture
talksApr 5, 2026
Speaking themes are now centered on AI delivery, sovereignty, and architecture
My talks and workshops are centered on practical AI adoption, agentic delivery, and architecture trade-offs.
talksaiarchitecture