Mapping Import
Immediate Predictive Enhancement
Once imported, the Semantic Match Engine proposes mappings for the rest of the target schema, lineage derivation activates, and impact analysis runs — extending the value of work your team has already done. Every model call writes a SemanticSuggestionReturned entry to the ledger.
There is no warm-up period. The moment a `MetaModelProjected` entry lands, the imported mapping is eligible for Predictive Mapping suggestions on adjacent target attributes. A spreadsheet that mapped 60% of the target schema before import becomes the seed corpus for the Semantic Match Engine to propose completions for the remaining 40%.
Each suggestion is generated by the `semantic-match` edge function (Google Gemini 2.5 Flash via the Lovable AI Gateway, temperature 0.1, normalized into one of five algorithm classes). Every model call writes `SemanticSuggestionReturned` carrying the model identifier, threshold profile version, confidence score, and the ULID of the source attribute it was generated from. Every accept or reject writes the corresponding ledger entry.
Lineage edges derive automatically from the freshly imported mappings. Each `MappingImported` row produces one or more `LineageEdgeDerived` entries — typically one per source-to-target field pair, more when transformations fan in or out. The lineage explorer becomes useful immediately, instead of waiting for a separate lineage scan to complete.
Impact Analysis becomes available on the first imported mapping. A proposed change to a target schema can compute its blast radius against the just-imported mapping graph, which is the most concrete payoff of moving a spreadsheet into the governed MetaModel: the spreadsheet could not answer 'what depends on this column?' without manual grep; the MetaModel can.
The same enrichment that makes Predictive Mapping work on imported spreadsheets also feeds the export side. When a CI/CD pipeline pulls the signed mapping spec to generate dbt YAML or Informatica XML (see ETL Integrations), the generated artifact carries the same parent-context and domain-vocabulary annotations the original spreadsheet only had implicitly.
All of this happens under the standard ledger discipline: SHA-256 hash-chained, role-at-time-of-action, draft-then-publish. There is no path through the Mapping Import workflow where Semantic Match suggestions, lineage edges, or impact reports bypass the ledger — which is what makes "predictive enhancement" a defensible audit story under 21 CFR Part 11 §11.10(e), HIPAA §164.312(b), and HHS Information Blocking review.
Audit ledger entries written by this capability
Each action in this capability appends one of the following entry types to the ULID‑keyed, SHA‑256 hash‑chained audit ledger. A reviewer can replay the chain to reconstruct exactly what happened, when, and under which role.
SemanticSuggestionReturnedSemanticSuggestionAcceptedLineageEdgeDerivedImpactAnalysisRunSee immediate predictive enhancement demonstrated against your data
Pick a sample dataset or bring your own. We will walk the ledger from ingest through Approver release in a single working session.
Request a demoRelated across the platform
Where this capability hands off — upstream adapters, downstream governance, and the underlying audit primitives.