Building the data and retrieval infrastructure behind an AI patent-analysis platform.

• Ingested 2.6M historical USPTO patents (1976–2001) into PostgreSQL, built their ArangoDB knowledge graph, and embedded them into a Qdrant vector index, extending searchable prior art back 26 years
• Migrated a 3.4M-record European patent corpus to PostgreSQL and backfilled 4.7M legal-status records plus 2.1M abstracts, raising coverage from 54.9% to 99.95% on an automated weekly sync
• Benchmarked 10 AWS instance configurations for embedding workloads, cutting cost from $25K to $13K and runtime from 60 days to 24 days
• Resolved 74 tracked issues: 37 production bug fixes, 17 user-facing features, and 21 data-infrastructure migrations, each taken from root-cause analysis through deployment and verification

Sole technical implementer for CRM, integration, telephony, and AI infrastructure.

• Architected a 10-stage CRM lifecycle with 9 automated workflow rules and 2 custom REST API integration pipelines
• Replaced a vendor connector that silently created folders invisible to the API, eliminating a data-integrity defect in patient-record intake
• Built an AI operating system of 9 custom Claude agent skills on a self-hosted MCP server, plus a 105-tool AI development environment
• Deployed Twilio telephony with a sequential ring-cascade call flow and voicemail fallback

Building AI-powered admissions intelligence solutions for universities.

• Built n8n-based web scraping and data enrichment pipeline producing 5,000+ university leads
• Defined workflows, KPIs, and success metrics for Tassl AI product development
• Aligned cross-functional inputs and shipped incremental releases

Founded and led a game development studio from concept to multiple releases.

• Led 12-person cross-functional team across development, design, and audio
• Ran 2-week Scrum sprints delivering 5 game initiatives
• Built business plan with rNPV valuation using 1,000+ Monte Carlo simulations
• Improved go/no-go decision accuracy by 35%

Delivered data-driven optimization of internal engineering and quality tools.

• Increased Diesel Needle Detector capacity from 10% to 60%
• Improved graph generation speed by 5× and database performance by 12.67×
• Migrated data to Apache Parquet enabling Power BI integration
• Built automated reporting pipelines reducing manual effort

Supported quality and process improvement through analytics and automation.

• Conducted Gage R&R study improving measurement accuracy by 82%
• Automated supplier email workflows for 234 suppliers using KNIME