Capacity optimisation — 50 seats
Removed 50 seats from a high-volume operation without service level impact. Data-driven, demand-aligned, structurally sound.
Operations and transformation. Twenty years of consequential work.

I build systems that perform — and people who own them.
Most operations problems look like process problems. They're usually people problems. Most AI problems look like technology problems. They're usually deployment problems. Twenty years in banking and BPO operations taught me to find the structural cause, not the visible symptom.
Inside a compliance-driven financial services operation, I led the programmes that actually changed how the work ran — capacity freed without touching service levels, an AI assistant built for frontline decision support, performance frameworks that created ownership rather than measurement.
Removed 50 seats from a high-volume operation without service level impact. Data-driven, demand-aligned, structurally sound.
Built an AI assistant that supports frontline decisions in real time. Handle time down. First contact resolution up. A human in the loop by design.
14 FTE saved through cross-skill enablement and macro automation. Process redesign that took the manual effort out at task level.
Frameworks that created accountability at every level — not just measurement of outcomes, but ownership of them.
Hiring, structured training design, coaching, performance recovery. The team that owns its outcomes is the intervention that compounds.
The structural decisions that make a business perform consistently — not the quarterly fix, the durable change.
How AI is actually deployed in environments where getting it wrong has a visible cost. Human oversight is not a limitation. It is an architectural decision.
How people grow inside systems — individually and in relation to each other. The culture that creates ownership is never accidental.
The preference for structural cause over visible symptom. The intervention that looks good in a quarterly review and the one that compounds over three years are rarely the same thing.
Markets. Sport. The patterns in how consequential decisions get made. Binary outcomes and clear signals — the same thing that makes operations interesting makes other domains interesting too.
Outside work, I run a personal automation workshop called Arizzgoat. It is where I build, test and learn: a research agent that proposes content angles and tags every claim as fact, inference or assumption; a drafting studio that writes only after I approve a direction; a review step where nothing goes out without my sign-off; and routine audits of how it all performs. It is a hobbyist’s bench. What works gets documented, and so does what fails.
If something I've written resonated, or you want to talk through anything in the areas above — send me a message on WhatsApp or connect on LinkedIn.