AI Automation · Workflow Engineering · RPA
TripleTen AI Automation graduate building production-style workflows that take repetitive, time-draining work off people's plates. Seven shipped projects spanning conversational AI (RAG), no-code orchestration, RPA, predictive automation, and conversion-focused web design.
Zapier · Make · n8n · UiPath · VAPI · Glean · Google Gemini · Claude & Claude Skills · OpenAI · Prompt Engineering · RAG · Document Understanding · Google Workspace APIs · Lovable
An end-to-end support workflow designed in Miro — not a shipped product. The diagram maps how AI could handle routine tickets end-to-end so CSRs focus only on novel cases, with tickets unable to close until the customer confirms resolution.
Live, publicly accessible chatbot grounded on a single WWII academic chapter that returns answers with citation-ready Chicago Manual of Style references.
Customer survey responses are automatically classified and summarized by Gemini 2.5, logged to Sheets, and route negative feedback straight to the right team via Gmail.
Eliminated manual morning planning for a weather-dependent kiosk; the workflow delivers staffing recs, product prep guidance, and ready-to-post social copy each day.
End-to-end RPA bot replaces manual invoice entry: extracts due date, supplier, total, invoice number, and description into a structured Sheet for every document.
Personal automation that turns my own delivery history into per-store ETA forecasts so I sequence stops around when trucks actually arrive.
Two production-quality conversion sites for local services, each engineered around a single business goal: turn visitors into booked appointments.
Live, explainable bracket predictor that finished within 2 points of the Sokol/LRMC academic benchmark (~75%), with every pick traceable to the underlying weighted factors. The title-game prediction (Michigan over UConn) was correct; three of four Final Four teams were in the formula's top 10.
A working, installed decision support system used on real personal decisions. The Brief turns vague questions into well-formed prompts; The Council runs them through nine thinking lenses with peer review and a chairman synthesis. Validated across three real decisions, with the chairman's confidence rating tracking real uncertainty rather than projecting false certainty.
This page is print-optimized. Use Print for browser-native PDF, or Download PDF for a generated single-page document.