Work that held up in production.

Challenge — Ageing MPLS links and flat networks caused branch outages and made compliance segmentation impossible.

What we did — Delivered an SD-WAN overlay with segmented VLANs, redundant links and centralised monitoring, rolled out branch by branch with zero unplanned downtime.

Challenge — Teams shipped manually into unmanaged accounts, with no consistent guardrails or audit trail.

What we did — Built a Terraform landing zone with policy guardrails, environment isolation and GitOps pipelines that took releases from monthly to daily.

Challenge — Claims officers spent hours cross-referencing policy documents, creating a persistent backlog.

What we did — Shipped a retrieval-grounded assistant with citation-backed answers and human review gates, integrated directly into the claims workspace.

Challenge — Alerts were scattered across tools with no ownership, and incidents were discovered by end users.

What we did — Consolidated telemetry into a single SIEM, wrote detection content mapped to MITRE ATT&CK and stood up an on-call response runbook.

Challenge — Purchase approvals moved by email and spreadsheet, with no visibility into where a request was stuck.

What we did — Automated the procure-to-pay flow with integrated approvals, ERP sync and a live audit trail for every request.

Challenge — Every region reported different numbers because each pulled from its own extract.

What we did — Rebuilt ingestion into a governed lakehouse with tested models and one certified metrics layer for the whole group.