Foundations
CMD+RVL runs on a modern data platform you already recognize—Airflow, dbt, Snowflake, catalogs, and observability—extended with richer metadata, lineage, and semantic access so market context stays explainable as data, assumptions, and outcomes change.
A familiar stack, extended where most stacks stop
CMD+RVL is intentionally built on standard data infrastructure patterns. The difference is that metadata, lineage, and context continue past ingestion and warehousing—connecting raw sources to derived fields, analytical views, and downstream outcomes.
EDGAR filings preserved as source truth.
All EDGAR filings are ingested in full into a raw data lake (Postgres + S3) before transformation. Original source artifacts are retained so revisions, restatements, and late filings can be reprocessed deterministically without relying on third-party normalization.
See Data ProductsAirflow and dbt with context-aware propagation.
ELT pipelines are orchestrated with Airflow and expressed in dbt. Models land in Snowflake as governed datasets, with transformations explicitly tied back to filing state and source context so teams can see not just that data changed, but why.
Explore EngagementsSnowflake-native datasets published as products.
Snowflake is the canonical warehouse. Data products are published through Snowflake Marketplace and other marketplaces, and the same datasets are used internally for Signals and engagements—ensuring one semantic baseline across delivery paths.
Browse Snowflake MarketplaceAnalysis designed for review and reuse.
CMD+RVL uses count.co as the primary analytics and communication layer. Views are treated as durable artifacts, linked back to datasets, assumptions, and event timing so conclusions survive sharing, review, and time.
Explore SignalsMetadata as the control plane
CMD+RVL treats metadata as a first-class system, not an afterthought. Operational metadata, semantic context, and outcomes are connected so teams can reason about trust, freshness, and impact without manual reconstruction.
Catalog-driven governance.
CMD+RVL maintains its own data catalog, built as a fork of Amundsen. The catalog includes ingestors, ownership, schema tracking, Slack alerts, and operational metadata, with APIs for programmatic read and write access.Extended lineage and assurance metadata.
The metadata model is extended beyond tables and columns to capture lineage, coverage, freshness, latency, schema history, and methodology notes—carried through to derived views and outcomes rather than stopping at the warehouse.Built to integrate, designed to last
Because CMD+RVL is grounded in standard data stack patterns—Airflow, dbt, Snowflake, catalogs, graphs, and observability—it fits naturally into existing research and production environments. The differentiation is how far context and lineage are carried.