Governance must produce evidence, not theatre
Policies matter only when they become ownership, control points, exceptions, approvals, monitoring, and auditable evidence inside delivery.
Expertise map
A practical architecture map for enterprise data, AI, cloud, risk, delivery, and leadership.
My work sits where strategy becomes operating architecture: governed data platforms, credible AI, hybrid-cloud foundations, delivery systems, and decision evidence that can stand up to executive, engineering, security, risk, and audit scrutiny.
Systems view: connect enterprise strategy, data estate, AI patterns, governance, cloud, and delivery into one coherent operating model.
Proven context: 25+ years across Swiss/DACH regulated environments, platform ownership, practice building, and executive delivery responsibility.
Working standard: make trade-offs explicit, keep architecture modular, and turn risk, lineage, controls, and decisions into reusable evidence.
Professional manifesto
The work is not to admire complexity. It is to make it governable.
My professional territory is organized around principles, questions, and capability areas that help enterprise data and AI become reliable operating systems.
Data OS beats platform sprawl
The future is not one more warehouse, lake, lakehouse, or catalog. It is a sovereign control plane across heterogeneous data and AI estates.
AI architecture is an operating discipline
Models are components. The real architecture includes data boundaries, evaluations, tool permissions, human review, cost, release gates, and rollback.
Enterprise architecture should reduce ambiguity
Good architecture turns stakeholder tension into explicit trade-offs, documented decisions, reusable patterns, and delivery behavior.
node 01
Enterprise Architecture
Target-state architecture, standards, portfolio rationalisation, transition roadmaps, and governance that make enterprise change safer, faster, and auditable.
Methods: Architecture principles and standards, Reference architectures and paved-road patterns, Architecture Decision Records, Capability and dependency mapping, Application lifecycle and technical debt scoring, ARB / Tech Council operating cadence.
Operated enterprise architecture governance forums and advised CIO/CDO, security, audit, and risk stakeholders across Swiss and DACH financial institutions.
Questions this answers
- What should the target-state architecture look like?
- Which platforms and applications should be rationalised first?
- How do we make architecture decisions auditable?
node 02
Data & AI Architecture
Enterprise data platforms, AI/GenAI guardrails, lineage, quality, and architecture patterns that move AI from experiments into governed production.
Methods: Enterprise data strategy, Data governance and quality, Data lineage and controls as evidence, AI/GenAI governance, Model risk liaison, LLMOps and AI-supported SDLC.
Built the Swiss/DACH Data Engineering & ML practice from 0 to 60+ FTE and delivered regulated data platforms and AI governance work.
Questions this answers
- Is our data foundation ready for AI?
- Where should AI controls enter the SDLC?
- How do we prove lineage, quality, and accountability?
node 03
Governance, Risk & Evidence
Regulatory evidence, BCBS 239 alignment, audit remediation, security patterns, and controls that help risk, audit, and delivery teams defend decisions.
Methods: BCBS 239 remediation, Audit findings remediation, Security and operational risk patterns, Compliance evidence packs, Exception handling, Controls as evidence.
Led regulatory-aligned initiatives where architecture, data lineage, audit remediation, and controls had to hold up under scrutiny.
Questions this answers
- Can we defend this system to audit and risk stakeholders?
- Which controls need evidence, ownership, and cadence?
- What is the minimum governance model that still works?
node 04
Hybrid Cloud & Platforms
On-prem plus Azure architecture, identity, Databricks/Delta Lake, IaC, CI/CD, and guardrails that create reusable platform lanes for regulated delivery.
Methods: Hybrid cloud architecture, Azure platform services, Identity and access patterns, Databricks and Delta Lake, Terraform / IaC, CI/CD and observability.
Architected regulated platform and integration patterns on hybrid cloud aligned with infrastructure, delivery, and governance constraints.
Questions this answers
- Which workloads belong on-prem, cloud, or hybrid?
- How do platform guardrails reduce delivery risk?
- How should identity, network, and data boundaries be designed?
node 05
Engineering Delivery
Delivery governance, SDLC quality gates, automation, Python/SQL/PySpark engineering, and observability that reduce delivery risk and production friction.
Methods: Delivery governance, Quality gates, AI-supported SDLC, Python, SQL, PySpark, Migration factories, Production observability.
Established delivery standards and quality gates while running multi-team transformation programmes in regulated environments.
Questions this answers
- How do we turn architecture intent into delivery behavior?
- Which quality gates matter before production?
- How do we scale migration and modernisation work?
node 06
Leadership & Operating Model
Practice building, P&L ownership, executive stakeholder management, team design, and operating models that improve delivery speed without losing control.
Methods: Practice building, P&L ownership, Executive stakeholder management, Team and role design, Mentoring architects and tech leads, Product/agile-at-scale operating models.
Owned a CHF 20M+ portfolio, sustained >85% utilisation, improved margin by >11%, and built teams across Switzerland and DACH.
Questions this answers
- What team shape is needed for enterprise data and AI?
- Which decisions should sit with architects, product, security, or risk?
- How do we build leadership cadence without bureaucracy?
Technical anchors
Tools matter because they define what can be operated, governed, automated, and changed.
Cloud, Data & AI Platforms
Platform choices and guardrails for governed enterprise data and AI delivery across hybrid and cloud-native environments.
Microsoft AzureDatabricksDelta LakeSnowflakeKafkaMLflow
Engineering, Automation & Delivery
Hands-on engineering credibility and delivery automation used to turn target architecture into repeatable production behavior.
PythonSQLPySparkTerraformCI/CDObservability
Architecture, Governance & AI Assurance
Decision systems and controls that keep enterprise architecture, data, and AI explainable, governable, and aligned with business value.
Data MeshLakehouseMLOps / LLMOpsData LineageBCBS 239Architecture Decision Records
Detailed capability briefings