Database production

From a database change to a release you can explain.

You produce the reference data that others rely on. Use VoLCA's inspection, editing, calculation and export capabilities within your own validation, review and publication process.

Diagram: Control, Modify, Recalculate, Review, Publish, Reference, version, Sources & edits, Expert review.
From a change to a reviewable release.An example cycle, connected to the organisation’s own review process.

One production cycle, connected building blocks

Start with a reference version. Make one deliberate change, examine its consequences, and prepare the material needed for review and release.

01 · Control

Check incoming data before integrating it

Inspect records, unresolved links, units and characterization coverage. Keep findings attached to the input version and the rules used.

Available building blocks: import, structural inspection and mapping diagnostics. Read the mapping audit guide and linked-database inspection example.

Integration to scope: run these checks in your CI and add your own acceptance rules. A successful import alone does not establish methodological validity.

02 · Modify

Correct a process without losing the starting point

Work on a writable copy, change a specific exchange and retain the source material and edit journal. Keep the reason for the correction with your review record.

Available building blocks: database editing and preserving source data and edited state.

Integration to scope: reviewer assignments, approval rules and your organization-specific change report. An edit journal is not by itself a complete collaborative review system.

03 · Recalculate

Understand the consequences across the database

Compute an agreed activity panel or the whole database with a pinned method and mapping configuration. Compare before and after, and examine failures as well as changed scores.

Available building blocks: batch scoring in the Ecobalyse export example and comparison against independent results in the BAFU numerical check.

Integration to scope: full-database coverage, change thresholds and release-blocking checks for your database. The published examples establish their stated workload, not automatic coverage of every database and method.

04 · Review

Give experts the material to make a decision

Bring together the changed records, inventory or impact differences, unresolved findings and calculation settings. Let reviewers inspect the evidence rather than only a final score.

Available building blocks: inspectable activity records, API outputs and the CSV and HTML comparison report example.

Integration to scope: connect these outputs to your existing review tools, access rules and approval process. Expert validation remains your responsibility.

05 · Publish and share

Deliver a version that others can identify and reuse

Export the approved data, retain the configuration and attach release metadata, checksums and validation reports. Share only what the data licences and your access policy permit.

Available building blocks: the checked multi-format export example and API access for downstream tools.

Integration to scope: your release catalogue, publication approvals, stable version addresses and sharing permissions. The export check demonstrates structural checks, not full round-trip equivalence or a ready-made institutional publishing platform.

Start with one real operation

Choose a correction or release check you already perform. Define its inputs, the expected report and the conditions that should stop publication. That is a useful first pilot without replacing your entire workflow.

Bring

An authorized sample or public database, one concrete change, your current validation rules and a reference result if available.

Check together

Can the operation be reproduced? Are all requested records accounted for? Can an expert explain the differences? Can your team retain the complete working state?

Evaluate the operating model

Review integration and data boundaries, maintenance and continuity, then scope hosting, access and support for your organization.

Examples using public Agribalyse or BAFU data are technical demonstrations, not endorsements by their publishers.