I design and modernize complex systems where enterprise architecture, cloud platforms and AI agents meet.
My work ranges from large-scale application modernization and cloud-native architecture to agent governance, observability, evaluation, identity, provenance and AI-assisted software delivery.
This GitHub is where I turn architecture ideas into executable evidence: reference architectures, PoCs, methods, demos, books and working code.
Architecture should not stop at the diagram. It should be testable, traceable and usable by the teams that have to deliver it.
| Area | What I am working on |
|---|---|
| Enterprise & Solution Architecture | Modernization strategies, architecture governance, platform boundaries, integration, identity and security |
| Agentic AI | Agent harnesses, MCP, multi-agent systems, ontologies, evaluation, drift, provenance and runtime governance |
| AI-assisted SDLC | AURORA-IA, intent/spec-driven development, Gherkin contracts, traceability and CI guardians |
| Azure & cloud-native | Microsoft Foundry, APIM, AKS, Dapr, Entra ID, OpenTelemetry, Application Insights and .NET 10 |
| Project | What it demonstrates |
|---|---|
| AURORA-IA — Animal, Vegetable, Mineral | A practical implementation of AURORA-IA with living specifications, Gherkin contracts and traceable AI-assisted delivery |
| COBOL → .NET 10 + React | Legacy modernization using AURORA-IA, Clean Architecture, CQRS and contract-driven migration |
| Agent Harness + CodeAct with .NET | Separation of agent control plane and execution plane, approvals, telemetry and tool-intensive reasoning |
| VS Code Custom Agents — OTel & Drift Audit | OpenTelemetry-based attribution, agent versioning, tool-call tracing and drift detection for GitHub Copilot custom agents |
| Governance Blueprints for AI Agents | Enterprise patterns for agent identity, audit, cost, safety, evaluation and source-control attribution |
| Ontology-Governed AI Agents | Deterministic business-rule enforcement using OWL ontologies and reasoning outside the LLM |
| TOGAF® 10 en Práctica | A pragmatic guide to applying TOGAF 10, including ArcDoc+ and AI as a transversal architecture capability |
AURORA-IA — AI-Unified Requirements, Orchestration, Reasoning & Automation is my approach to software delivery in the age of AI agents.
Its core idea is simple:
Intent → Living Specification → Behavioral Contracts → Code → Verification → Traceability
The objective is not to generate more code. It is to make AI-assisted delivery governable, verifiable and repeatable.
A recurring theme in my work is that enterprise agents need more than prompts and models.
I explore practical controls for:
Identity · Mandates · Provenance · Observability · Evaluation · Drift · Human approval · Cost attribution · Source-control attribution
The goal is to make agentic systems operable under the same engineering and governance expectations we already apply to enterprise software.
I work across the full path from enterprise intent to executable architecture:
- Legacy modernization and decomposition
- Application and enterprise architecture
- Cloud-native and distributed systems
- Identity and zero-trust integration patterns
- Architecture governance and ADRs
- Platform engineering and delivery guardrails
- AI-assisted engineering and agentic workflows
My default approach is evidence-driven: understand the real system, make the trade-offs explicit, build a thin proof, measure it, and then scale.
Azure · Microsoft Foundry · GitHub Copilot · .NET · C# · Kubernetes · Dapr · APIM · Entra ID · OpenTelemetry · Application Insights · Service Bus · SQL Server · PostgreSQL · Terraform · Bicep · MCP
- jmfloreszazo.com — architecture, AI, modernization, Azure and experiments
- TOGAF® 10 en Práctica — free practical architecture book
- SpeakerDeck — presentations and technical material
- Sessionize — speaking profile
- YouTube — demos, ideas and fragments of what I am building