Data & Security Analytics Consultant
I help organizations make sense of their data and defend the systems that produce it by building the dashboards, models, and monitoring workflows that turn raw information into decisions people can act on with confidence.
About
I work at the intersection of analytics and security operations. I've learnt and built reporting pipelines for retail and finance teams, modeled churn and credit risk, and spent time on the monitoring side triaging alerts and correlating logs for security operations center. That combination shapes how I approach a new dataset, I want to know not just what the numbers show, but how the system producing them could fail or be misused.
Most of my work lives in SQL and Python, with Power BI or Tableau for the reporting layer. I've spent time inside SIEM tooling as well, which has made me more careful about data lineage and access than I might otherwise be. I like projects with a clear business question attached, a churn number a retention team can act on, a risk score a lender can defend, an alert threshold a SOC team can actually live with.
Selected work
Organized by industry. Each project reflects a real business question, the constraints around the data, and what was delivered at the end.
Built a scoring model to flag loan applications with elevated default risk before approval.
Business problem
The lending team was approving loans on limited historical context, resulting in an unpredictable default rate quarter to quarter.
Tools
Deliverables
Reviewed transaction logs to identify unusual patterns worth escalating to the fraud team.
Business problem
Manual review of flagged transactions was falling behind volume, and thresholds hadn't been revisited in over a year.
Tools
Deliverables
Identified the customer segments most likely to cancel and what typically preceded it.
Business problem
The retention team had no early signal for at-risk customers and was mostly reacting after cancellations happened.
Tools
Deliverables
Forecasted inventory needs across categories ahead of peak seasonal demand.
Business problem
Overstock and stockouts were both happening in the same quarter, in different regions, for the same product lines.
Tools
Deliverables
Modeled the likelihood of 30-day readmission using intake and discharge records.
Business problem
Readmission rates affected reimbursement, and the care team had no ranked list of patients to follow up with first.
Tools
Deliverables
Tracked bed occupancy and staff allocation across departments in one view.
Business problem
Department leads were pulling utilization numbers from separate systems that rarely agreed with each other.
Tools
Deliverables
Set up cohort tracking to see which new features actually changed retention.
Business problem
The product team was shipping features without a clear read on whether they moved any retention metric.
Tools
Deliverables
Designed and analyzed a test comparing two onboarding sequences.
Business problem
Drop-off in the first session was high, and the team had two competing ideas for fixing it with no data to settle it.
Tools
Deliverables
Rebuilt triage priority rules to cut down on repetitive low-value alerts.
Business problem
Analysts were spending most of a shift on alerts that turned out to be benign, while genuine incidents sat in the queue.
Tools
Deliverables
Documented indicators and response steps for a recurring phishing pattern targeting staff.
Business problem
Response to phishing reports varied by analyst, and there was no shared reference for what to check first.
Tools
Deliverables
Reworked delivery routing to reduce distance and idle time across regional depots.
Business problem
Routes had been set manually years earlier and hadn't adjusted as new store locations were added.
Tools
Deliverables
Traced where perishable stock was being written off and why.
Business problem
Write-off costs were rising, but nobody had broken the losses down by supplier, route, or storage location.
Tools
Deliverables