Back to Resume

Selected Work

Project Case Studies

Systems and automation case studies spanning intelligence pipelines, AI-enabled prospecting workflows, translation automation, recommendation modeling, and a recruiter automation CLI built with TypeScript and Node.js.

Medisca · Ongoing initiative

Competitive Intelligence Pipeline

Designed and built an event-driven competitive intelligence system that monitors competitor activity and delivers decision-ready briefs to GTM teams.

Problem

Competitive updates were shared manually by the commercial team, creating inconsistent visibility across the organization.

Technical approach

  • Architected event-based web scraping pipeline to monitor competitor content, pricing, and launch signals.
  • Integrated Claude for automated analysis, converting raw page changes into structured intelligence briefs.
  • Built automated alert distribution delivering real-time competitive updates to Sales, Product, and Marketing.
  • Designed internal dashboard giving GTM teams self-serve access to current competitive intelligence.

Stack & systems

  • Python
  • Firecrawl
  • Claude
  • n8n

Outcomes

  • Eliminated manual competitive monitoring across three departments.
  • Reduced time from competitor signal to team awareness from days to hours.

Medisca · Ongoing initiative

AI-Enabled Sales Prospecting Pipeline

Built an AI-enabled prospecting system using Clay for data enrichment and Claude for lead scoring, automating list generation, qualification, and sales handoff.

Problem

Outreach and list building were manual and inconsistent across territories.

Technical approach

  • Built automated data enrichment pipeline using Clay to pull firmographic and licensing data from multiple sources.
  • Integrated Claude for AI-powered lead scoring and qualification against ideal customer profile criteria.
  • Automated detection of newly registered license holders and surfaced them as sales-ready opportunities.
  • Designed recurring alert system delivering enriched prospect lists segmented by territory and priority.

Stack & systems

  • Python
  • Clay
  • Claude
  • Google Business Profile API

Outcomes

  • Reduced manual list-building effort from days to minutes per territory.
  • Accelerated new-account discovery and improved sales handoff quality.

Medisca · Ongoing initiative (Communications + Design)

Translation Workflow Automation

Architected and deployed an end-to-end automated translation system that reduced a five-day manual cycle to near-instant delivery, saving $500K+ annually.

Problem

Manual translation took around five days: documents were sent to a platform, quoted, accepted, and then queued for delivery.

Technical approach

  • Designed user-facing upload interface requiring zero training - users drop files and receive translated outputs automatically.
  • Triggered translation jobs directly from file uploads in a SharePoint upload folder.
  • Automated translation processing with DeepL through Power Automate while preserving formatting and styling.
  • Applied glossary and consistency controls so translated outputs stay aligned with channel language standards.
  • Routed translated drafts to the channel owner for QA and approval, then returned approved files to a separate folder.

Stack & systems

  • SharePoint
  • Power Automate
  • DeepL

Outcomes

  • Increased design-team throughput by 40%.
  • Increased communications-team throughput by 20%.
  • Delivered over $500K in savings in the first year.

Medisca · Ongoing optimization initiative

Product Recommendation Model

Scoped, designed, and delivered a product recommendation engine driving $1M+ in incremental revenue through improved cross-sell performance.

Problem

Product discovery was low on the website experience.

Technical approach

  • Defined recommendation logic using purchase behavior and add-to-cart signals as primary inputs.
  • Implemented recommendation modules directly on product detail pages.
  • Returned product recommendations to increase discovery depth for active shoppers.
  • Built GA4 attribution model to measure recommendation-driven revenue against baseline.
  • Measured and optimized performance using GA4 attribution modeling.

Stack & systems

  • Algolia
  • GA4
  • A/B testing

Outcomes

  • Drove over $1M in revenue impact through recommendation-led product discovery.

Personal project · Active development

Recruitment Automation CLI

Designed and built an end-to-end automation pipeline that scrapes ATS platforms, parses email alerts, scores jobs with AI, and saves results to Notion.

Problem

Manual job searching across dozens of company career pages, email alerts, and job boards was time-consuming and inconsistent.

Technical approach

  • Built TypeScript/Node.js scrapers for Greenhouse, Lever, and Ashby ATS platforms, pulling structured job data from 200+ companies.
  • Integrated Gmail (IMAP) parsing to process LinkedIn and Indeed job alert emails and extract structured listings.
  • Designed AI-powered job scoring using Gemini, matching resume keywords, skills, and location preferences against job descriptions.
  • Built Notion integration to deduplicate and store analyzed jobs with match scores, skills gaps, and red flags.
  • Implemented automated email digest delivering top matches with AI analysis summaries.

Stack & systems

  • TypeScript
  • Node.js
  • Gemini AI
  • Notion API
  • Gmail (IMAP)

Outcomes

  • Processes 200+ company career pages and email alerts in a single automated run.
  • Reduces manual job search from hours to minutes with AI-scored priority ranking.