Document Parsing Rules

Docparser applies a rule the customer has already set up, such as a template matched to a specific invoice layout, to pull named fields and tables out of an incoming PDF or scanned document. A Beam agent watches the inbox or folder Docparser monitors, waits for a new document to be parsed, and reads the resulting fields once Docparser has applied the rule. It then passes those values into the destination system, for example an accounting tool, without re-keying anything by hand. Where Docparser cannot match the document to a known rule, the agent leaves it unparsed for a person to set up.

Line Item Extraction

Beyond single fields, Docparser can pull repeating rows, such as invoice line items or order quantities, out of a document as a structured table. A Beam agent reads that table once parsing finishes and writes each row into the matching record in the destination system, for example splitting an invoice into individual expense lines. It keeps the source document linked to the resulting entries so a person can trace any figure back to the original file. When row counts look inconsistent with the document, for example a total that does not match the summed lines, the agent holds the record for review.

Multi-Format Document Intake

Docparser accepts documents arriving by email, upload, or a connected storage folder, so a Beam agent does not need a single fixed intake point. The agent watches whichever channel the customer has configured, picks up a document as soon as Docparser finishes parsing it, and applies the same downstream steps regardless of how the document arrived. This keeps handling consistent whether an invoice came from a supplier's email or a scanned upload. If a document arrives in a format Docparser does not support, or arrives through an unconfigured channel, the agent notifies a person instead of dropping it silently.