VP, Data + Information Management

Position Summary The VP, Data & Information Management is the architect and owner of Pave America's enterprise data layer, master data management practice, and the data foundation that makes applied AI and ML possible at scale. This is a greenfield build at a PE platform company spanning three deliberately balanced disciplines: (1) Master Data Management - the canonical record of customers, vendors, jobs, assets, employees, and chart of accounts across all acquired brands, with the golden-record / merge-survivorship logic that turns all disparate source systems into one trustworthy enterprise view; (2) Data Modeling & Semantic Architecture - a Snowflake-based Enterprise Data Warehouse built on disciplined dimensional modeling, a Kimball-style mart layer, governed dbt project architecture, a semantic layer in Power BI (or equivalent), and the Direct Margin (DM) formula codified consistently across PavementSoft and NetSuite; (3) Applied AI / ML Data Platform - the active ownership of feature stores, label pipelines, training data quality, and ML evaluation infrastructure that VP Innovation's ML team consumes. The VP, Data is not the consumer of AI requirements - the role is the active builder of the data foundation that determines whether AI at Pave succeeds or stalls. Outputs feed branch-level P&L, operational KPIs, M&A integration, AI use cases, and AEA board reporting from a single source of truth. Key Responsibilities Master Data Management: Own the MDM practice end-to-end across all brand entities - golden-record design and merge-survivorship logic for customer, vendor, job, asset, employee, and chart-of-account dimensions; MDM tool selection (Reltio / Profisee / Informatica MDM / Tibco EBX or the right-sized alternative for Pave); data stewardship operating model with named stewards in Finance, Ops, and HR; the MDM-vs-EDW boundary; data quality scoring and remediation cadence; and the explicit role MDM plays in M&A integration (Day-1 customer dedup, vendor consolidation, asset onboarding). Data Modeling & Semantic Architecture: Lead the dimensional modeling and semantic architecture of the EDW - staging / intermediate / mart layering inside dbt; Kimball-style fact and conformed-dimension design; slowly-changing-dimension handling (SCD Type 1/2/4 patterns); explicit data contracts between source systems and the warehouse; naming conventions, model documentation, and lineage discipline; semantic-layer governance in Power BI or equivalent that prevents metric drift across consumers. This is the discipline that determines whether the EDW scales gracefully or rots over five years. Direct Margin (DM) formula standardization: Own the end-to-end Direct Margin formula across PavementSoft and NetSuite; partner with the CFO and the NetSuite implementation partner to enforce a single canonical DM definition; encode it in the semantic layer and reconcile the variance to <1% between source systems. EDW design & build: Lead the Snowflake (or Redshift) + dbt + Power BI + reverse ETL stack build, managing the SI partner. Stand up source feeds from PavementSoft, NetSuite, HubSpot, Paycor, Samsara, and Limble with documented data contracts and validated pipelines. Applied AI / ML Data Platform: Actively own the data foundation that makes ML possible - feature stores, label pipelines, training data quality, point-in-time correctness for time-series features, ground-truth instrumentation, and ML evaluation infrastructure. Establish the clean hand-off boundary with VP Innovation: VP Data owns the data foundation (features, labels, training sets, evaluation pipelines, monitoring); VP Innovation owns the models (training, tuning, deployment, productization). KPI framework: Deliver the VCP-linked KPI framework (revenue, DM, EBITDA, crew productivity, bid-to-actual variance, cash conversion, working capital) with clear ownership, refresh cadence, and a single canonical definition encoded in the semantic layer - not in a hundred Excel files. BI & s…