Interface
Multi-department interface, server and client rendering depending on each screen's real need.
Architecture & Evidence
I am not limited to the interface: I also design the functional, technical and governance architecture that makes the solution possible.
Data flows, cloud, applied AI, security, governance and deployment — the depth behind the demonstration.
From business need to system, never the other way around.
1. Functional Architecture
Every layer has a clear responsibility — operational systems stay the source of truth, integration never duplicates authority, and AI only steps in once the data is reliable.
- 1Operational systems (ERP, on-prem SQL Server, department systems)
- 2Integration layer (events, synchronization)
- 3Consolidated data (BigQuery)
- 4Business services
- 5Multi-department interface
- 6AI-assisted human decisions
A stack chosen layer by layer, not stacked out of habit.
2. Technical Architecture
Data
BigQuery as the consolidated analytical source of truth; on-prem SQL Server stays the operational system of record — no duplicated authority.
Cloud & integration
Serverless deployment, file storage, asynchronous events, scheduled jobs, secrets never hardcoded.
Applied AI
AI-assisted extraction and reasoning — always paired with deterministic validation, never shipped raw.
Security & identity
Federated authentication, role-based permissions — never implicit access to sensitive data.
CI/CD
Continuous integration and deployment, versioned and traceable images before anything reaches production.
One truth, never a duplicate.
3. Data
The question isn't "where does the data live" but "who owns it." BigQuery consolidates for analysis and cross-department visibility; existing operational systems stay the authority on their domain.
Architecture decision example — illustrative
- Context
- The on-prem SQL Server system already owns a department's operational data.
- Decision
- On-prem SQL Server stays the system of record for that domain.
- Solution
- The consolidated platform stores references and analytical projections in BigQuery, not an authoritative copy.
- Why
- Avoid dual authority, protect governance already in place, reduce migration risk.
- Rejected
- Creating a parallel database that would duplicate the operational data.
Portability and cost where they earn their place.
4. Cloud
Cloud Run means paying for actual usage instead of always-on capacity — relevant for variable operational load. Pub/Sub decouples components so a slowdown in one service doesn't block another; Secret Manager keeps credentials out of the code.
Useful, never uncontrolled.
5. Applied AI
Document AI extracts information from incoming documents; Vertex AI / Gemini reasons over the consolidated context. In both cases, deterministic validation protects critical business rules — the same principle as the previous page: AI proposes, it never executes an operationally impactful action alone.
Identity and permissions as a foundation, not an afterthought.
6. Security
Okta/OIDC for federated authentication instead of a homegrown password system. RBAC so every role sees exactly what it needs — never more. Secrets live in Secret Manager, never in code or a versioned config file.
Traceability and accountability, by design.
7. Governance
Every data domain has a clear owner. Important architecture decisions are documented, not just implemented — so the next person understands why, not just what.
Ship often, without breaking what works.
8. CI/CD
- 1Commit and code review
- 2Automated build (Azure DevOps)
- 3Tests
- 4Versioned image (Artifact Registry)
- 5Deployment (Cloud Run)
- 6Post-deployment observability
What he actually did, not what he claims he can do.
9. Steve's Actual Role
| Domain | Contribution |
|---|---|
| Functional and technical architecture | End-to-end design, from business need to implementation. |
| Data and cloud | Chose BigQuery as the analytical source of truth; Cloud Run / GCS / Pub-Sub architecture. |
| Applied AI | Integrated Document AI and Vertex AI / Gemini, with deterministic validation. |
| Security and governance | OIDC authentication, RBAC permissions, secrets management. |
| CI/CD | Azure DevOps pipelines, versioned images via Artifact Registry. |
The kind of mandate this profile serves best.
10. Target Roles
Operations
Director of Operations
Plant Director
Multi-site Director
Director of Maintenance / Reliability
Director of Operational Excellence
Director of Continuous Improvement
Digital Transformation
Director of Digital Transformation
Director of Business Solutions
Director of Business Systems
AI & Automation Manager
Director of Automation
Director of Innovation
Business / Technology Intersection
Digital Transformation Advisor
Business Architect
Senior Transformation Advisor
Business-oriented Solution Architect
Business Solutions
Product Operations
20+ completed trainings on Google Cloud Skills Boost.
Continuing Education
Between March and June 2026, structured training covering cloud infrastructure, security, networking, Kubernetes, Terraform and generative AI — all completed and passed.
Training completed
Generative AI Leader Certification
Completed March 2026 — Google Cloud
Generative AI and agents
- Gen AI: Beyond the Chatbot
- Gen AI: Unlock Foundational Concepts
- Gen AI: Navigate the Landscape
- Gen AI Apps: Transform Your Work
- Gen AI Agents: Transform Your Organization
- Engineer AI Agents with Agent Development Kit (ADK)
Architecture and infrastructure
- Google Cloud Fundamentals: Core Infrastructure
- Essential Google Cloud Infrastructure: Core Services
- Elastic Google Cloud Infrastructure: Scaling and Automation
- Reliable Google Cloud Infrastructure: Design and Process
- Cloud Architecture: Design, Implement, and Manage
Networking and security
- Networking in Google Cloud: Fundamentals
- Networking in Google Cloud: Network Architecture
- Implement Cloud Security Fundamentals on Google Cloud
- Managing Security in Google Cloud
Deployment and automation
- Manage Kubernetes in Google Cloud
- Develop Serverless Applications on Cloud Run
- Getting Started with Terraform for Google Cloud
- Implementing Cloud Load Balancing for Compute Engine
- Set Up an App Dev Environment on Google Cloud
Certifications in progress
- Professional Cloud Architect Certification
- Professional Data Engineer Certification
- Cloud Architect Accelerated Path for AWS professionals
- Integrate Generative AI Into Your Data Workflow
Next step
Let's discuss your business problem.
Describe the current situation — a real operational problem is enough of a starting point.