Open to CX systems and automation work
Eight years in CX operations, five of them shipping production automations. Most recently that meant owning Zendesk end to end for roughly 150 agents at Robinhood, and writing the Python and SQL that kept the rest from turning into manual work.
agents on the Zendesk instance I administered, across four teams
RobinhoodSLA adherence on internal application issues
Robinhoodlower cost per ticket after new internal tooling shipped
Upsideminutes for the weekly provisioning run, once it was scripted
Robinhoodmore one-touch resolutions with account context in Zendesk
UpsideA little context
I sit between the support floor and the codebase.
That combination is the whole point. I have worked the queue, so I know which manual step is actually worth automating — and I can write the thing, take it through testing and deployment, monitor it, and leave a runbook behind instead of handing it off at launch. From Louisville, now based in Chicago.
Where I have been
Support rep → tooling manager, 2017—26Customer Support Representative
Beltsville, MDTechnical support across the product suite in Salesforce and Zendesk, plus ownership of at-risk accounts.
Customer Support Team Lead
Tysons Corner, VABuilt the Zendesk–Jira integration and introduced a deflection chatbot. Onboarded the team on SQL and Zendesk.
Operations Specialist
Washington, DCShipped internal tooling in JavaScript, put customer data in front of agents, and wrote the 50+ articles behind a 15% deflection lift.
Tooling Manager
Chicago, ILOwned Zendesk end to end for ~150 agents across four teams, and led the support organization’s migration to Salesforce.
Systems I have built
06 case studiesInternal application issues had no intake path of their own, and not all of them were Zendesk work.
Built the system that takes them in and routes them, then held 95% SLA adherence against it.
Reg E disputes and CCPA and GDPR data subject requests carry deadlines the organization has to meet and then prove it met. Tracked by hand across teams, that fails quietly — you find out you missed a window afterward.
Automations linking Zendesk to internal systems that measure adherence as the work happens and surface a request before it runs past its window, so compliance is something reported rather than assumed.
On a ~150-agent desk, two agents would routinely pick up the same issue in parallel.
A scheduled service that detects duplicates and merges them through the REST API, taking hours of repeated work off agents and administrators every week.
The support organization had to change platforms without losing history, continuity, or response times.
Led object and field mapping, data migration, and cutover planning across every affected team. SLA came through the cutover unchanged.
Tickets that moved between teams broke per-team SLA numbers, so staffing decisions were running on bad data.
A model of transfer-based ticket ownership that produces accurate team- and agent-level metrics for staffing and prioritization.
A weekly account provisioning and outreach run cost an hour of manual work every time it came around.
Scripted end to end, with monitoring and a runbook. The same run now finishes in under five minutes.
On my own time
Turns Apple Reminders into real calendar blocks using a deterministic, unit-tested scheduling engine. No backend, no accounts, no network calls.
Repository private while it is in progressWhat I work in