Skip to main content

Field Resolution

Field resolution is the process that matches each extracted field against Talonic's Field Registry using a three-band matching model. The bands determine whether a match is accepted automatically, flagged for confirmation, or treated as a new field. Resolution is the core mechanism that turns raw, document-specific field names into canonical registry entries, building a unified knowledge graph across all your documents.

ParameterTypeDescription
Auto band≥ 0.80 similarityHigh-confidence match. The field is linked to the existing registry entry and its occurrence count is incremented.
Confirm band0.50 – 0.79Candidate match. The field is linked but flagged for manual review in the Review Queue.
New band< 0.50No match found. A new Provisional field and cluster are created in the registry.

How Resolution Works

Resolution follows a strict three-band order that is never skipped. First, the system checks for an exact name match against existing registry entries. If no exact match is found, it checks for a cluster member match — whether the field name matches any synonym in an existing semantic cluster. Finally, it computes semantic embedding similarity using AI embeddings to find conceptually similar fields. This graduated approach prioritizes fast, deterministic matches before falling back to more expensive similarity comparisons.

Resolution runs concurrently across documents. Each document's fields are resolved in an isolated transaction to prevent lock contention, and occurrence counts are updated atomically in the same SQL transaction using upserts with a 3-attempt deadlock retry mechanism. This keeps the registry eventually consistent without blocking concurrent ingestion, even when hundreds of documents are being processed simultaneously.

After resolution completes, the platform evaluates maturity promotions, regenerates affected schemas, and updates the cross-schema harmonization view in a fixed chain: resolve, then promote, then regenerate. The chain is never interrupted, so newly promoted fields immediately appear in auto-generated schemas. The resolution process also feeds into the job pipeline: during Phase 1 of a job run, the system uses a 3-tier lookup cascade (string normalization, token fuzzy matching, then AI fallback) to fill 60-80% of cells without a full LLM call, dramatically reducing cost.

Pending field decisions from the confirm band flow into the central Review queue (/review). Accept to merge into an existing cluster, or reject to create a new field.

Monitoring Resolution via API

Batch resolution runs automatically as documents are processed, so there is nothing to schedule or trigger. To monitor it, use the read-only registry health endpoint, which includes occurrence and recent-resolution statistics alongside tier counts and the deterministic transfer hit rate. It is the same view the in-app registry dashboard reads, so a script polling it sees exactly what you see in the app.

Check registry and resolution health
curl https://api.talonic.com/v1/registry/health \
  -H "Authorization: Bearer $TALONIC_API_KEY"

For best results, process pending confirmations promptly in the Review queue. Unconfirmed fields in the confirm band (0.50-0.79 similarity) remain in a pending state that can affect downstream extraction accuracy. Confirming correct matches strengthens the cluster and improves future resolution, while rejecting incorrect matches prevents bad data from propagating through the knowledge graph. Teams that review confirmations weekly typically see their auto-band match rate increase steadily over the first few months of platform usage.

Frequently asked questions

How does field resolution work in Talonic?+
Each extracted field is matched against the registry using three bands in strict order: exact name match, cluster member match, then semantic embedding similarity. Results fall into auto (>=0.80, auto-linked), confirm (0.50-0.79, flagged for review), or new (<0.50, creates a new Provisional field). The three-band order is never skipped.
Where can I review pending field confirmations?+
Pending field decisions appear in the central Review queue at /review. Accept to merge the field into an existing cluster, or reject to create a new independent field. Processing these decisions promptly improves resolution accuracy for future documents.
What happens after resolution completes?+
After resolution, the platform evaluates maturity promotions and regenerates affected schemas in a fixed chain: resolve, then promote, then regenerate. This ensures that newly promoted fields immediately appear in auto-generated schemas. The chain is atomic — it never breaks midway.
How does resolution reduce extraction cost during job runs?+
During job runs, the system uses a 3-tier lookup cascade — string normalization, token fuzzy matching, then AI fallback — to fill 60-80% of cells without a full LLM call. Fields that are well-established in the registry with high occurrence counts are the most likely to resolve via lookup, making Canonical and Established fields essentially free to extract.