Role. Security Analyst, SecureU project, University of Connecticut (Feb 2012 - Oct 2013)

Problem

A large public university runs thousands of staff and faculty machines, and over years of normal work those machines accumulate sensitive data they were never meant to keep: Social Security numbers in old spreadsheets, scans of forms, exported reports, email attachments saved to a desktop and forgotten. Every one of those files is a breach waiting to happen. The SecureU project existed to find that Personally Identifiable Information across university computers and get it remediated. The hard part was not running a scan once. It was doing it at institutional scale, on machines owned by people who had real work to do, without breaking their workflows or losing their cooperation.

Approach

  • Used Identity Finder to scan machines for PII, SSNs, financial account data, and other regulated identifiers, then triaged each hit into shred, redact, or securely relocate, rather than blindly deleting files people might still need.
  • Stood up Splunk indexing to aggregate scan results across many machines, so remediation could be tracked and prioritized centrally instead of one desktop at a time.
  • Worked day to day in Linux for the tooling and scripting side, and built and maintained the project's web presence in HTML, CSS, and PHP so departments had a clear place to understand the process and report status.
  • Treated this as a behavior-change problem, not just a cleanup job: sat with the people whose machines held the data, explained why a file was a liability, and walked them through remediation so the same data did not pile back up after I left.
  • Handled regulated data carefully and consistently, since a sloppy remediation that leaks the very SSN it was meant to remove is worse than no scan at all.

Impact

  • Reduced the university's exposed-PII footprint by clearing sensitive data off machines across departments, removing files that were genuine breach risks.
  • Gave the program a real audit trail: with results indexed in Splunk, remediation status was visible and trackable instead of living in scattered scan logs.
  • Shifted habits, not just files. Walking owners through why each file mattered meant fewer repeat offenders and a cleaner baseline over time.
  • Early, direct exposure to security work, regulated data, Linux tooling, and the human side of getting people to change how they handle information, that still shapes how I think about guardrails and data handling on AI systems today.

Stack

Identity Finder (PII discovery and remediation) · Splunk (result indexing and aggregation) · Linux · HTML · CSS · PHP

Details are generalized to respect my employer's confidentiality; the specifics are ones I can stand behind in an interview.