Docparser

@Docparser list all parsers in my account and show me which ones have processed documents in the last week

Your account has 4 active parsers: 1. **Invoice Parser** (ID: inv_2024) — 47 documents processed this week 2. **Receipt Scanner** (ID: rcpt_main) — 12 documents processed this week 3. **Contract Extractor** (ID: contract_v2) — 0 documents processed this week 4. **Purchase Order Parser** (ID: po_legacy) — 3 documents processed this week The Invoice Parser is your most active, handling vendor invoices with fields for amount, date, and line items. The Contract Extractor hasn't seen activity recently—consider archiving it if no longer needed.

This example uses the List Parsers tool to show account-level visibility. Docparser doesn't natively return "last week" metrics in the list call—the AI would need to follow up with Get Multiple Document Results per parser to count recent documents. Useful for auditing which parsers are actually in use versus sitting idle.

@Docparser fetch the last 20 parsed results from my Invoice Parser and calculate the total amount billed this month

Retrieved 20 invoices from your Invoice Parser (ID: inv_2024). Here's the breakdown: **Total billed in March 2024:** $18,347.52 Top vendors: - Acme Supplies: $6,200.00 (4 invoices) - Widget Co: $4,850.00 (3 invoices) - Office Depot: $2,100.00 (5 invoices) - Misc vendors: $5,197.52 (8 invoices) All invoices were dated between March 1-15. Note: This reflects the last 20 documents processed, not necessarily all March invoices if you've processed more than 20 this month.

This example combines Get Multiple Document Results with AI reasoning to sum extracted fields. Docparser returns structured JSON per document—the AI parses the 'amount' field and aggregates. Requires that your parser template already extracts an 'amount' field; if field names vary, the AI may struggle to identify the right data without guidance.

@Docparser show me the model layouts for my Contract Extractor parser and explain when I'd use each one

Your Contract Extractor parser has 3 model layouts configured: 1. **Standard Contract** — Default layout for single-page agreements. Extracts party names, effective date, term length, and signature block. Use for NDAs, consulting agreements, or simple service contracts. 2. **Multi-Page MSA** — Layout for Master Service Agreements spanning 5+ pages. Adds extraction zones for payment terms, liability caps, and renewal clauses. Use when processing enterprise vendor contracts. 3. **Amendment Layout** — Specialized for contract amendments. Focuses on amendment date, original contract reference, and modified clauses. Use when vendors send change orders or addendums. Docparser auto-selects the best layout based on document structure, but you can manually assign layouts via the dashboard if auto-detection fails.

This example uses Get Parser Models to surface layout configurations—helpful when a parser handles document variations (invoices with/without line items, contracts with different page counts). The AI explains the *purpose* of each layout, which Docparser's API doesn't provide—that context comes from layout names and field lists. Useful for onboarding teammates or debugging why a document parsed incorrectly.

When Docparser beats manual invoice entry for small finance teams

A 3-person accounting team at a SaaS company receives 40-60 vendor invoices per month as PDFs via email. They need line items, totals, and vendor details in their ERP before month-end close. Docparser wins here if the invoices follow 3-5 consistent templates—its parser models handle layout variations without retraining. The Get Multiple Document Results tool pulls batches of parsed data into Switchy, where the team reviews flagged fields before pushing to QuickBooks. If your invoices are one-offs from random vendors with wildly different formats, Docparser's accuracy drops below 85% and manual review time erases the gain. For repeating vendor relationships, this setup cuts invoice entry from 90 minutes to 15 minutes per batch.

Why Docparser struggles with high-volume resume parsing

A 6-person recruiting team at a Series A startup screens 200 resumes per open role. They want to extract skills, work history, and education into an Airtable base for candidate ranking. Docparser is borderline here—resumes lack the structural consistency of invoices, so parser accuracy hovers around 70% even with multiple models. The team spends more time fixing extraction errors than they save on manual data entry. If you're hiring for 2-3 roles per quarter with 50 resumes each, the setup cost might break even. Beyond that threshold, a resume-specific parser or an LLM-based extraction tool delivers better accuracy without the model-tuning overhead. Docparser's 4-tool scope is too narrow for this workflow's edge cases.

When Docparser fits legal teams with template contracts

A 2-person legal ops team at a B2B company manages 30 customer contracts per quarter, all generated from the same MSA template with rider variations. They need renewal dates, liability caps, and payment terms in a Notion database for proactive account management. Docparser is the right call if 80% of contracts share the same clause structure—the parser models lock onto section headers and table layouts. The Get Parser Models tool lets the team handle 2-3 rider variations without rebuilding parsers from scratch. If your contracts are heavily redlined or negotiated from scratch, Docparser's fixed-layout assumption breaks and you'll spend hours tweaking models per document. For template-driven contracts under 50 pages, this setup extracts clauses in under 2 minutes per file.