The normalization infrastructure enterprises have  been missing

Rosetta Stone Attributes give you a catalog of canonical, standardized definitions—email, age, product ID, country, etc—that sit as a semantic layer above raw supplier schemas. Query attributes as a first-class source, preserve lineage through transformations and materialized views, and join consistently across every supplier without rebuilding bespoke mappings for each new dataset or partner.

AI generates field-level mappings at query time, eliminating the standard ingestion pipelines you typically travel before data becomes usable. When schema mapping hits the limit, row-level ML and LLM-powered classification handles the long tail automatically. You manage mappings through a full lifecycle, and in Snowflake-native deployments, inference runs close to the data plane, returning suggested mappings as metadata without touching underlying data.

Normalization is the hidden time-tax that kills collaboration. It’s slow, fickle, and repeats endlessly. AI-assisted mapping turns it into an ongoing, scalable capability so new data becomes usable faster and stays usable as schemas drift further apart.

Graph-agnostic identity resolution supports configurable deterministic and probabilistic matching to fit your use case, with BYO logic to override for custom match logic when you need it. This architecture lets you swap identity partners without destabilizing the rest of your stack. Because Rosetta Stone normalizes identity-related attributes upstream, your resolution layer operates on cleaner, standardized inputs which improves match quality and consistency across every source.

Incoming data is normalized on arrival; then governed, tracked, and mapped to Rosetta Stone Attributes so it's immediately queryable alongside your own datasets.Access rules, licensing terms, rate cards, and row-level provenance travel with every transaction, giving you a governed collaboration surface instead of an unstructured file exchange. Source exactly what you need, join it immediately, and skip the normalization sprint.

It creates liquidity and speed. Source the exact raw data you need, avoid overbuying, and shorten time-to-value. Because datasets are interoperable as soon as you access them.

Data collaboration fails without trust. Governance ensures compliance, prevents accidental over-sharing, enables safe internal/external collaboration, and makes monetization viable because licensing/permissions are enforceable and auditable.

• Save query outputs directly as access rules, attach permissions, pricing, and tags, then reuse them safely across workflows 
• Collaboration Policies define exactly what each party can view, map, and share
• Governance enforces at the row level while provenance and policy implications travel with every output for clean auditing and downstream enforcement
• Apply opt-out and privacy filters at query time in the data plane so excluded identifiers never surface in results

NQL is SQL-like by design. It's tailored to Narrative's abstraction model, querying Rosetta Stone Attributes as a source, referencing access rules by name, and joining across governed external datasets without custom pipelines. NQL supports interactive query previews—so you can validate logic before committing—and incremental view maintenance for efficient large-dataset refreshes.

Pair it with Query Builder or Data Studio for teams that need a UI layer. Or stay in the editor and work at full power. NQL makes cross-supplier data feel like it lives in one logical database and accelerates iteration for both technical and semi-technical users.

Run normalization and collaboration workloads directly inside your cloud environment. Compute comes to the data, not the other way around. In Snowflake-native deployments, in-environment components pull sample data locally, run mapping and normalization inference, and return results and metadata without moving underlying data. Keep your governance consistent, simplify compliance, and eliminate ingestion pipelines. All without ceding control of your data plane.