Helen of Troy
Progressed from frontline support into team supervision, with work spanning coaching, QA, productivity reporting, and customer feedback analysis.
AI ENABLEMENT & WORKFLOW AUTOMATION
I build AI workflows around the way teams work.
I spent nine years at Helen of Troy, progressing from frontline support to supervising teams of up to 20. Today I combine that operations experience with AI-assisted research and workflow automation in paid contract work.
Open to AI enablement, AI operations, and CX automation roles. Based in Hopedale, Massachusetts, and open to hybrid or on-site work in the area.
EXPERIENCE
Operations judgment, applied to working systems.
Roadshow Concepts
Support municipal licensing assignments using AI-assisted research, reusable SOPs, document organization, and a workflow tracker. I verify official sources and personally approve outgoing outreach.
Selected work
03 PRIMARY PROJECTS + 01 ADDITIONALLive demo
AI QA Scorer: transcript scoring and coaching
- Problem
- Support supervisors need a faster way to find rubric gaps, check the evidence, and prepare useful coaching without giving up the final decision.
- My contribution
- I designed the rubric-based review flow and built the public demonstration around synthetic transcripts, visible evidence, and editable coaching drafts.
- Evidence and status
- This demo draws on an earlier QA workflow I used at Helen of Troy, where I estimate it reduced overall manual QA work by about 40%. The public demo has not been benchmarked for time savings.
Technical details
React · OpenAI Responses API · fixed rubric scoring · structured output
Live demo
ReviewSignal: customer review trend intelligence
- Problem
- Customer-review patterns can reveal questions worth investigating before they become clear in support contacts alone.
- My contribution
- I built a dashboard that groups prepared review examples into themes, keeps source comments visible, and turns a pattern into a question for evidence review.
- Evidence and status
- The dashboard uses synthetic data. It recreates an analysis process I used at Helen of Troy, where review signals were investigated and shared with marketing and engineering rather than treated as proven causes.
Technical details
React · TypeScript · synthetic review data · evidence review
Case study + demo
AI support email triage with human approval
- Problem
- A small consumer-products support team is buried in repetitive order-status, product, refund, and complaint emails.
- My contribution
- I built the routing logic, reply-draft path, human approval step, test integrations, audit logs, workflow downloads, and proposed pilot plan.
- Evidence and status
- A connected test version exercises Gmail, Slack, Jira, and Sheets with synthetic messages. Fixed demo rules were used for the connected pass. No production inbox, customer reply, adoption result, or measured time saving is claimed.
Technical details
n8n · OpenAI API · Gmail · Slack · Jira · Google Sheets
Live AI demo
SupportOps: one policy set for chat and MCP
- Problem
- Most support chatbots answer confidently whether or not they actually know. This project gives people a cited chat UI and gives AI clients the same synthetic policy set through a tool-scoped MCP server.
- My contribution
- I built the policy retrieval, cited answer view, bounded tool set, stored demo handoff, and four-case evaluation.
- Evidence and status
- The demo uses a synthetic policy library and keeps uncertain cases visible for human review. The MCP download includes three read-only tools and one write tool limited to a local demo escalation.
Technical details
React · OpenAI Responses API · MCP · source citations · D1
ABOUT PETER
Support leader who builds working systems.
I spent nine years at Helen of Troy, moving from frontline consumer support into supervision. I directly supervised teams of up to 20 in an operation of about 40 agents. My work included coaching, QA, customer feedback analysis, and a daily productivity-reporting workflow that I estimate cut manual work for that specific report by about 95%. Today I use AI-assisted research in paid licensing assignments, with official-source checks and my approval before any external message.
01
Team leadership
Led frontline support teams and coached people through live customer work.
02
Applied AI
I use AI-assisted research and workflow tools in current paid contract work.
03
Quality and coaching
Turns support standards into clear rubrics, review steps, and useful coaching.
04
Working prototypes
Builds small tools people can test, inspect, and improve.
Want AI your team will actually use?
Open to AI enablement, AI operations, CX automation, and support systems roles. Based in Hopedale, MA (Boston MetroWest), and open to hybrid or on-site roles in the area.