If you scroll through enough tech portfolios or LinkedIn profiles, you’ll see a lot of talk - strategy slides, vision decks, capability matrices. But rarely do you see the messy, hands-dirty work of actually fixing real problems.

That’s where I’ve spent most of my time.

Working across the energy, government, healthcare, and finance sectors in Australia and New Zealand, I’ve consistently been called in when the diagram stopped working - or when the platform everyone signed off on didn’t deliver. This isn’t a case study post. This is a reflection on real-world tech issues, and how I’ve helped teams solve them.

1. Data Silos Are Killing Insight

What I found: Business teams couldn’t access meaningful insights because data was trapped in silos - across spreadsheets, on-prem systems, and legacy APIs.

What I delivered:

• Unified streaming + batch pipelines using Azure Data Factory and Kafka

• Designed data platforms with Unity Catalog + Databricks for governance

• Enabled self-serve data models that made reporting instant, not painful

2. Cloud Spend Out of Control

What I found: Cloud bills were ballooning - teams were over-provisioning, and nobody owned optimisation.

What I delivered:

• Introduced FinOps practices (budgets, tagging, rightsizing workflows)

• Migrated heavy compute to serverless and event-driven architectures

• Helped align architecture with business outcomes - not just lift and shift

3. Multi-Cloud Security Was a Patchwork

What I found: Inconsistent security practices across AWS, Azure, and SaaS tools, leading to shadow IT and unmanaged risks.

What I delivered:

• Built a Zero Trust Architecture with proper identity federation

• Embedded DevSecOps into pipelines with automatic security scanning

• Set up compliance workflows that actually passed audits without stress

4. AI Projects with No Guardrails

What I found: Teams wanted AI. But they lacked governance, ethics, and often, a clear purpose.

What I delivered:

• Developed AI strategy roadmaps focused on business value, not novelty

• Implemented MLOps with model monitoring, bias checks, and human-in-loop

• Used GPT and NLP where it actually saved time - not just to tick a trend box

5. Legacy Systems, Everyone’s Elephant in the Room

What I found: Everyone feared touching the old systems - but they were holding everything back.

What I delivered:

• Used Strangler Fig pattern to modernise gradually via microservices

• Defined domain-driven designs to simplify and segment core capabilities

• Reduced risk and cost by evolving, not rewriting, what still worked

6. Digital Identity Is Still Broken

What I found: Identity and access were inconsistent - especially across public-sector platforms.

What I delivered:

• Explored Decentralised Identity (DID) and Verifiable Credentials

• Developed citizen-focused flows with secure, wallet-based access

• Aligned builds to NZ Privacy Act and Australia’s Digital ID framework

7. Sustainability Goals, But No IT Metrics

What I found: Organisations had ESG targets but couldn’t measure IT’s contribution or carbon impact.

What I delivered:

• Embedded GreenOps into cloud architecture reviews

• Integrated carbon tracking into data workflows and dashboards

• Enabled ESG reporting through blockchain traceability and IoT

8. IoT and GIS in Utilities Were Disconnected

What I found: Field asset data from SCADA, IoT, and GIS systems wasn’t integrated - making dashboards unreliable and delayed.

What I delivered:

• Connected real-time streams with Kafka, TimescaleDB, and Databricks

• Enabled predictive analytics for power quality and asset maintenance

• Built digital twins for network modelling and load forecasting

9. DevOps Was Just a Buzzword

What I found: Agile ceremonies were in place - but deployments still happened over email or remote desktop.

What I delivered:

• Set up GitOps workflows, blue-green deployments, and rollback plans

• Enabled observability-first pipelines using Prometheus and ELK

• Fostered a shared Dev+Ops culture, not just DevOps tooling