Rational Data Services is a Houston-based boutique consultancy for modern cloud data platforms. We design and modernize Data Lakes, Lakehouse, and deterministic Data Fabric architectures on Databricks, Snowflake, and Trino for federal, financial, healthcare, higher-education, and commercial clients — bringing big-firm rigor and senior architect–led delivery.
Governed, platform-native data products — from ingestion through the agentic layer.
End-to-end design and delivery on Databricks and Snowflake.
Expose governed data to AI agents through Model Context Protocol and a semantic layer, so every query resolves deterministically and in context.
Fine-grained, attribute-based policy with OpenMetadata, Apache Ranger, OPA & Trino.
Streaming, event-driven, and declarative pipelines across Bronze, Silver, and Gold.
Modernization across AWS, Azure, OCI, and Snowflake / Databricks.
Natural-language query via Databricks Genie, Snowflake Cortex, and curated semantic layers.
The rigor of IBM, KPMG, Guidehouse, and Lockheed Martin — without the overhead drag.
Engagements led by a 20+ year practitioner, not a bench.
IRS, Federal Reserve, CDC, VA, IMF & national security.
Agents answer from real governed data — every result reproducible and traceable, never a hallucination.
Deep Databricks, Snowflake & Trino specialization, not generalist.
From data strategy through pipeline build and agentic enablement.
Senior delivery with no overhead drag and direct access to the architect.
Designed and implemented a deterministic Data Fabric — distinct from a lakehouse — that lets analysts and AI agents ask complex logistics, planning, and readiness questions in natural language. Engineered the governed data layer beneath an Agentic / MCP interface, enforcing strict Need-to-Know access through attribute-based policy: OpenMetadata for cataloging and lineage, Apache Ranger and Open Policy Agent (OPA) for real-time policy enforcement, and Trino for federated, deterministic query across siloed sources — so every agent answer is auditable and respects clearance and NTK boundaries.
A near-real-time Databricks Lakehouse integrating IoT signals (location, viewer counts, ad displays) from a mobile ad fleet. Medallion layers built with Lakeflow Declarative Pipelines power dynamic pricing, fleet utilization, and self-service analytics.
A Snowflake Data Lake unifying Gem (ATS), NetSuite, and adjacent SaaS to expose bottlenecks and efficiency metrics, with modern ELT via Fivetran and dbt. A Snowflake Cortex chatbot lets recruiting and ops teams query it in natural language.
A consolidated Identity & Data Access Management repository spanning Student, Applicant, and Donor SaaS systems for a major NYC university, delivered as a Medallion Architecture with Informatica and dbt.
Data intake, governance, and analytics use-case framework for a multi-tenant enterprise data platform on Databricks — Bronze-layer pipelines and a correlated data model enabling analyst-driven exploration across business users.
Whether you're modernizing a platform, standing up governance, or preparing your data for agentic AI — let's talk about what senior, architect-led delivery can do for your team.