Integrations, exports and API: what goes in, what comes out.

Sources with a purpose-built reader

These formats have their own reader in Planlined. Everything else goes through the general reader backed by Google Cloud Document AI, then the same verification screen where every value sits beside the page it came from.

• Medical carrierAetnaFunding Advantage and level-funded quotes
• Medical carrierUnitedHealthcareQuotes and monthly invoices
• Medical carrierFlorida BlueRate sheets and quotes
• Ancillary carriersDental, vision, life, STD and LTDProposals from Principal, Guardian, The Standard, Pacific Life, MetLife, Lincoln Financial, Unum, Cigna, Mutual of Omaha, Aflac, VSP, EyeMed and Delta Dental
• PEOADP TotalSourceProposals and renewal packets
• Payroll and HRISPaychex and AuraAura quote comparisons and HRIS output rate sheets carrying BlueCross Texas, Aetna, UnitedHealthcare, Cigna, Humana and Paychex MEC plans
• Quoting platformBenefitterMulti-carrier proposal exports
• Enrollment materialEmployee benefit guidesPDF and Word open enrollment guides with plan comparison grids
• Rate workbooksExcel and CSV rate gridsRead cell by cell, deterministically, with no model in the loop
• Carrier recognitionIncumbent plan namesCigna, Humana, Kaiser Permanente, Anthem, BlueCross BlueShield, Oscar Health, Molina Healthcare, Centene, Health Net, Ambetter and AvMed are identified from the plan names on incumbent and renewal documents
• Everything elseGeneral readerSBCs, quotes, rate sheets, invoices, census files and benefit summaries from any other carrier. Uncertain values are flagged, never guessed.

Carriers Planlined recognizes by name

The readers above are built for specific document formats. These carriers are recognized by name wherever they appear, which is what puts the right carrier card, phone number and provider-lookup link into the proposal book and the employee guide. Documents from a carrier without its own reader go through the general reader and land on the same verification screen.

Files under 1 MB parse while you wait. Larger files run as a job you can poll or receive a webhook for.

How it connects across your stack

A REST API at app.planlined.com with API keys, signed outbound webhooks, and connectors for the systems brokers already run. Available on Enterprise plans. The OpenAPI document is available on request.

• AuthenticationBearer tokens or API keys with scopesSign in tokens and Planlined OAuth access tokens as bearer, or an X-API-Key header generated in workspace settings. Keys carry scopes such as cases:read, cases:import, exports:write and pdf:extract. Partner apps can also obtain tokens through Planlined's own OAuth token endpoint, with secret rotation.
• WorkflowsOne capabilities endpoint, five workflowsGET /api/v1/integrations/capabilities returns the endpoints, scopes and retry policy for parse, case import, benefit guide publishing, proposal book PDF and proposal book workbook. Long jobs expose status, result and cancel. Requests accept an X-Idempotency-Key so a retried call cannot create a duplicate.
• ParsingSend a document, get structured plansPOST /api/v1/parse/unified runs the same reader registry the app uses. Files under 1 MB return synchronously; larger files return a job id to poll. Type-specific routes exist for census, quote, SBC, invoice, spreadsheet, image and XML.
• Case importPush a packet in, get a case or a proposed mergePOST /api/v1/case-imports retains every source file, parses all plans and benefit types, and either creates a case or proposes a version-guarded merge into an existing one. Automatic mode commits only when every benefit maps unambiguously; review mode waits for a person. Retained sources can be reprocessed later.
• ExportsArtifacts your system can fetchProposal book PDF and XLSX, hosted benefit guide links, JSON, and versioned artifact downloads. Render endpoints take the artifact content the app produces; they do not yet render from a bare case id.
• LimitsDocumented, not surprisingIntegration routes allow 60 requests a minute per key; webhook management routes allow 30. Files are capped at 50 MB. Synchronous parses have a 60 second budget before they become jobs.