AI leadership
Shape roadmaps, align stakeholders, guide cross-functional teams and make deliberate build, buy and scale decisions.
AI & Data Science Leadership
Strategy · Architecture · Delivery
Leading AI from strategy to production.
I bring together senior leadership, deep technical expertise and hands-on execution to turn complex data and AI opportunities into useful, governed enterprise systems.
StrategyTrustScaleDeliveryAI
LeadershipHead of Data Science
LeadershipDirector of Data Analytics
TechnicalLead Data Scientist
Current workPrincipal AI & Data Consultant
What I bring
Leadership that stays close to the technology.
The strongest AI programs connect business direction, technical judgment and disciplined delivery. My work spans all three.
Technical depth
Work across machine learning, LLMs, agentic AI, knowledge graphs, GraphRAG, causal reasoning and explainability.
Enterprise delivery
Move from problem framing and architecture through proof of concept, implementation, governance and operational adoption.
Technical foundation
Hands-on across the enterprise AI stack.
Current work combines AI architecture with implementation across models, data, cloud platforms, integration, evaluation and operations.
View the full technical stack →AI & agent engineering
Python · FastAPI · LangChain · LangGraph · LlamaIndex · Hugging Face
Machine learning & XAI
scikit-learn · PyTorch · TensorFlow · XGBoost · SHAP · LIME
Cloud, data & graphs
AWS · Azure · SQL · Neo4j · Snowflake · Databricks
Production & observability
Docker · CI/CD · LangFuse · OpenTelemetry · NeMo Guardrails
Selected project experience
Enterprise AI built around consequential decisions.
The emphasis is on systems that improve how organizations find evidence, assess risk and act with confidence.
View the delivery portfolio →Knowledge-grounded AI assistants
Architectures that combine governed organizational knowledge, semantic retrieval and graph traversal to produce contextual, traceable answers.
- GraphRAG and hybrid retrieval
- Evidence, permissions and provenance
- Agentic workflows with bounded tools
Regulatory and risk intelligence
Connected-data approaches that help analysts navigate counterparties, transactions, ownership, obligations and supporting evidence.
- Fraud and risk modeling
- Entity and relationship analysis
- Explainable investigation workflows
Document intelligence and agentic workflows
AI systems that retrieve, extract and reason over complex documents while preserving review points, sources and operational controls.
- Retrieval and structured extraction
- Human-in-the-loop workflow design
- Evaluation and guardrails
Predictive and explainable analytics
Models and decision-support systems designed for business contexts where performance, transparency and challengeability all matter.
- Predictive modeling and prioritization
- Causal and graph-based reasoning
- Model explanation and monitoring
Further work
Technical writing, speaking and selected advisory engagements.
These activities extend the central work: leading, designing and delivering AI. They provide additional ways to share ideas and support organizations.
Technical author
I write about AI, data science, knowledge graphs and enterprise delivery on Medium and LinkedIn, where I have nearly 33,000 followers. I am also developing a technical book, Knowledge Graphs for AI.
Explore technical insights →Speaker & trainer
Keynotes, seminars and technical sessions for leaders and practitioners working with AI, data and connected knowledge.
Invite Paulo to speak →Advisory engagements
Focused guidance on AI strategy, architecture, roadmaps, proofs of concept and difficult technical decisions.
Explore advisory work →Technical perspective
Architecture follows the question—not the fashion.
LLMs are one component of a broader system. Reliable enterprise AI also needs data quality, retrieval, permissions, evaluation, observability and people who can challenge the result.
Looking for leadership that can also deliver?
Let’s discuss an AI leadership role, enterprise project or difficult technical decision.