Vetta

Field staff photograph equipment nameplates, AI reads the values, and a person reviews and approves each record before it goes to the asset system.

The problem

Asset records are often incomplete because nameplate details are copied by hand in the field, then re-typed into the asset system. It is slow, and errors are hard to catch.

Vetta is designed around a simple rule: the AI proposes, a person decides. Every extracted value is shown next to the photo it came from, and nothing reaches the asset system until a reviewer approves it.

Built today

• A phone and tablet capture app that works offline in the field.
• Photo-quality checks before a picture is accepted.
• AI extraction of nameplate values, with every record reviewed and approved by a person.
• Matching against duplicates and existing equipment records.
• Export of approved records to an asset-system load workbook.
• Single sign-on through the organization's identity provider.

Current status

Vetta is in pre-production and available for controlled pilot evaluations to validate functionality and alignment with agency requirements.

• Not yet deployed for a customer.
• Extraction accuracy has not yet been measured on a formal test set, so we don't quote accuracy figures.