Data engineering professional with 25+ years building data foundations across six technology generations -- from mainframe ETL to modern cloud-native pipelines with dbt, Snowflake, and LLM-powered orchestration.
Currently building AI-powered tools and data systems independently:
I designed and operate a production-grade personal finance automation platform using Python, LangGraph, and LLM-powered validation. The system orchestrates a 19-node agentic pipeline ingesting from multiple financial data sources (brokerage APIs, budgeting platforms, rental property management, tax documents), with Gemini-powered checker nodes producing IRS-ready outputs across Schedules B, C, D, and E. Built with dbt (13 models, 58 tests), DuckDB, and MCP server integrations.
LoreConvo (github.com/labyrinth-analytics/loreconvo): An MCP server plugin for Claude that provides cross-surface persistent memory across Claude Code, Cowork, and Chat. Features 12 MCP tools, SQLite with full-text search, auto-save/load hooks, and a CLI interface. Built in Python under my Labyrinth Analytics practice.
I also operate a team of scheduled autonomous agents that handle opportunity research, architecture review, and implementation across my projects -- mirroring the multi-agent team patterns that enterprises are beginning to deploy.
My strengths: translating complex stakeholder needs into scalable data solutions, picking up new tools fast, and taking end-to-end ownership. I've consistently been brought in to build analytics functions from scratch -- most recently designing an enterprise data warehouse for a major cancer research institute, and previously scaling BI across three Disney divisions using Airflow.
Specialties: Data Engineering, Analytics Engineering, Agentic AI, Multi-Agent Systems, MCP Development, dbt, Python, SQL, Snowflake, Airflow, LangGraphMore...