AI Architect & AI Product Manager
Carlo
De Rossi
Helping organizations transform AI experimentation into governed, scalable business value.
25+ years building enterprise technology in regulated financial services.
What I bring
Areas of Expertise
Spanning AI strategy, engineering, and enterprise governance.
AI Strategy & Governance
Designing AI governance architectures for regulated environments. From policy to implementation under FINMA, GDPR, and the EU AI Act.
Agentic AI
Building production-grade agentic systems — multi-agent orchestration, tool protocols, and self-healing ML pipelines.
AI Product Management
Translating AI capability into business outcomes. Discovery, prioritization, and delivery across 700+ enterprise stakeholders.
Enterprise Architecture
End-to-end platform design at scale. CHF multi-million budget governance, 35K+ deployments, vendor-neutral principles.
Azure AI
Hands-on Azure AI engineering: MLOps pipelines, Azure ML Champion-Challenger, AI Foundry multi-agent deployments.
Program Leadership
Leading enterprise-scale programmes. SOX Key Process Control owner with zero audit findings across PwC and EY.
Career
Experience
25+ years building platforms and leading AI transformation in global financial services.
2023 – Present
UBS
·Director, Product Manager and Stream Lead
Zürich, Switzerland
Global Wealth Management Platforms- ›Leading AI Interdependency Stream — Analysed ~22,000 application relationships using graph analytics and ML (Louvain, SCC, GNN, HDBSCAN), reducing structural complexity by 53.1% and supporting retirement of 2,000+ applications
- ›Designing CIAA v2.0: FINMA/GDPR/EU AI Act compliant agentic framework with context-isolation and local-first inference
- ›AI adoption analytics across ~7,500 employees; graphRAG benchmarking (Neo4j vs ChromaDB) over EU regulatory corpus
- › Designed a custom risk assessment model for application decommissioning combining application complexity, business criticality, data sensitivity, compliance requirements and historical incidents; integrated the resulting risk score into a clustering model to minimise decoupling risk and cost.
- ›Acted as Agile Product Owner within Agile@UBS crews, driving roadmap prioritisation, product decisions and OKR alignment across cross-functional teams. Chapter Lead / Line Manager of an Engineering team.
2001 – 2023
Credit Suisse
·Product Manager, Senior Architect, Risk & Service Manager
Zürich, Switzerland
Enterprise Platforms · AI/ML · FinOps- ›Product Manager & Senior Architect of the Application Platform — 750+ global clients, 35K+ deployments. Technical Owner of the related technologies and products.
- ›Deputy Owner, Container Platform — Avoid Vendor Lock-In principle, 350-app migration from commercial to open-source
- ›SOX Key Process Control Owner — zero findings across PwC and EY audits across 22 years
- ›FinOps: Managed multi-million CHF platform economics and delivered $5M recurring annual cost avoidance while reducing vulnerability exposure from 80% to <10%
- ›Managed enterprise contracts exceeding USD 10M annually, supporting more than 750 internal customers and 25,000 platform deployments.
- ›Led a large-scale digital transformation program, migrating more than 350 enterprise applications and 35,000 deployments to an open-source platform, delivering over USD 25M in cost avoidance.
- ›Managed product strategy, lifecycle, licensing, chargeback, vulnerability management and vendor relationships. Partnered with business and technology stakeholders to design enterprise reporting, inventory and chargeback solutions.
- ›AI/ML applied to time-series forecasting, stakeholder clustering, vulnerability risk analysis and neural forecast models; 700+ stakeholders managed
- ›Applied AI/ML techniques, including neural forecasting models and graph-based clustering, to improve forecasting and operational decision-making.
- ›Led architecture and technical strategy for global Application platform.
- ›Provided architecture consulting for Online Banking, Trading and Private Banking systems.
- ›Defined enterprise Java technology standards
- ›Managed vendor relationships and contracts valued at approximately USD 12M annually
- ›Architect and Line Manager of an Engineering Consulting team
1998 – 2001
GFT Consulting
·Consultant
Zürich / Berlin
Forward-deployed to Swiss Post, Deutsche Post, ESTV- ›Systems integration and platform delivery for Swiss Post, Deutsche Post, and Swiss Federal Tax Administration (ESTV)
- ›Early experience bridging technical delivery with institutional client stakeholders in regulated public-sector environments
AI Architecture & Governance Thinking
Thought leadership on agentic AI, governance, and enterprise transformation.
AIQ Is Not Enough: Bainbridge 2.0 and the AI Autonomy Paradox
An exploration of the relationship between AI autonomy and human capability, examining how increasingly autonomous AI systems may change the skills, judgment, and capabilities required from the humans who design, supervise, and govern them.
AI AutonomyWhy AI Product Managers Need Hands-On Skills
An argument for AI Product Managers to develop practical technical capabilities alongside product management skills, including understanding AI models, agents, evaluation, APIs, data, and implementation constraints.
AI Product ManagementThe Next European AI Unicorn May Be a Trust Layer for Autonomous Agents
An exploration of the opportunity for a European AI trust layer that provides governance, identity, policy enforcement, observability, and accountability for autonomous AI agents operating across enterprise systems.
Agentic AIAI Agents Don't Need More Guardrails. They Need Governance.
An examination of why enterprise AI agents require governance across autonomy, identity, permissions, tool use, decision-making, accountability, and operational controls rather than relying only on technical guardrails.
AI GovernanceTETHER: A Governance Framework for Air-Gapped AI
A governance framework for AI systems operating in air-gapped and edge environments, addressing trusted attestation, episodic synchronization, task-bounded autonomy, human or hardware override, evidentiary logging, and rollback readiness.
AI GovernanceFrom Model Risk Management to Autonomous AI
An incremental governance framework extending traditional Model Risk Management from classical AI to Generative AI, RAG, Small Language Models, and autonomous agents, with increasing emphasis on behavioral assurance and decision provenance.
AI GovernanceGitHub
Projects
Production implementations across agentic AI, MLOps, RAG, and financial ML.
graphRAG
Graph-based Retrieval-Augmented Generation system.
GenAI / RAG / Graph MLmlflow-pipeline
End‑to‑end production MLOps pipeline using MLflow.
MLOps / Infrastructureself-healing-ml-agentic-demo
Agentic MLOps system with autonomous drift detection and retraining.
Agentic AI / MLOpsadaptive-rebalancing-agent
Adaptive portfolio rebalancing agent.
Financial ML / Agentic AICoMMEnTo
Agentic AI pattern implementation.
Agentic AI / Architectureauction-agents
Multi-agent RL auction simulation.
Reinforcement Learning / Multi-Agent Systemsa2a
A2A multi-agent communication protocol implementation.
Agentic AI / Protocolscarlo-career-mcp
Custom MCP server for career assistance.
AI Infrastructure / MCPNeuralForecast4TimeSeries
Neural forecasting models with Optuna optimization.
Time Series / Financial MLagentic-rag
Agentic RAG system with autonomous reasoning.
GenAI / Agentic RAGAIPdMML
AI Product Management + Machine Learning notebooks.
AI Product ManagementCreditCardFraudAnomalyDetection
Financial anomaly detection model.
Financial ML / Anomaly DetectionCredentials
Selected Credentials
A focused selection of credentials supporting AI architecture, agentic AI, governance, Azure, and enterprise transformation.
Executive profile
AI Architecture · AI Strategy · Governance · Enterprise Technology
Selected certifications are shown here for relevance to senior AI architecture and transformation roles. The complete credential inventory spans AI, machine learning, Azure, data, product management, architecture, governance and enterprise technology.
AI & Agentic AI
Current AI engineering, agentic AI and AI business architecture capabilities.
Governance & Architecture
Architecture, AI governance, risk and management-system disciplines.
Azure & Data
Microsoft Azure, data science and ML engineering capabilities.
Enterprise & Product
Enterprise technology management, service management and product disciplines.
UBS Internal Credentials
Enterprise credentials reflecting internal roles and capabilities developed within UBS.
Looking for a senior AI architecture, strategy, product, or transformation leader?
Discuss a role