Enterprise IT leadership.
Practical AI engineering.

The work I care about sits between engineering and operations: deciding what should be automated, building it so it can be operated, and making sure someone can tell what happened when it fails.

That perspective comes from years of production support and incident management in financial services — where changes are reviewed, failures have consequences, and documentation matters after the person who built something has moved on.

• Clear ownership and explicit scope before build
• Change control and evidence of what actually ran
• Monitoring, retry behaviour, and recovery designed in
• Documentation that outlives the original author