Founder and Principal Consultant
GovConDash
Novi, Michigan, United States
Designed, built, and operated GovConDash, an intelligence and compliance platform that transforms federal acquisition, contract, and legal data into actionable insights. The platform integrates federal data sources into a governed analytics environment with automated data pipelines and end-to-end agentic workflows.
- Executed full lifecycle data engineering across five federal data sources (eCFR, SAM.gov, USAspending, GAO, ASBCA, COFC, Federal Circuit, SBA, SEC EDGAR, DoW Comptroller budget exhibits, and govinfo.gov legislative text (NDAA)), automating ingestion into a PostgreSQL data warehouse with Apache Airflow orchestration, data quality gates, and disaster recovery through scheduled backup validation.
- Managed 127 production datasets by developing data dictionaries, governance standards, natural key relationships, classification taxonomies, and knowledge graphs for more than 18,000 unstructured legal decisions, enabling relationship discovery, pattern detection, and trend analysis across compliance data.
- Evaluated and integrated emerging AI, analytics, and data platforms including PostgreSQL/pgvector, LangChain, LangSmith, OpenAI models, MCP, Databricks, Docker, VS Code, and Power BI, validating adoption decisions through retrieval accuracy, latency, and cost analysis.
- Automated multi-day manual compliance research workflows by building 20+ production agentic workflows across six compliance domains, each executing a deterministic sequence that retrieves from governed corpora, cites every source, and emits a client-ready deliverable.
- Engineered verification-first AI governance by validating every AI-generated citation, quotation, and financial value against source evidence, while logging every model execution to an auditable trail with token-level cost monitoring.
- Analyzed ASBCA decisions and GAO bid protests to produce quarterly intelligence reports highlighting legal trends, emerging procurement risks, and actionable recommendations.
