The agent calls Docsumo's Post MCA Analysis endpoint with an array of document IDs representing the bank statements. Docsumo processes the documents and returns a month-by-month breakdown of account credits, debits, and balances. The agent then presents this financial summary to your team or the customer directly in the conversation.

Docsumo supports invoices, bank statements, checks, tax forms (including IRS forms), ACORD insurance forms, utility bills, debt settlement letters, purchase orders, and receipts. The platform uses OCR and machine learning to achieve 99% extraction accuracy across semi-structured document formats.

Tars uses your Docsumo API key, which you find under the Integrations section of your Docsumo account. The key is passed as an X-API-KEY header with every request. No OAuth setup or additional permissions configuration is required.

No. Tars calls Docsumo's API in real time during conversations. Extracted data from invoices, bank statements, and other documents is used only to generate responses. Tars does not maintain a separate database of your financial document data.

Yes. The agent calls Docsumo's User Document Types endpoint, which returns your current plan limits and usage. It can inform you or your team members about remaining page credits before initiating large batch uploads, preventing quota overruns.

Docsumo's IDP platform combines OCR with machine learning models trained on millions of document samples. It achieves 99% accuracy on printed text and handles common handwriting patterns. For poor-quality scans, Docsumo flags low-confidence fields for human review rather than guessing.

Docsumo's dashboard requires a team member to upload documents, wait for processing, and manually review results. With Tars, the AI agent handles document submission and result retrieval during live conversations. Customers and team members get financial data answers without logging into a separate platform.

Yes. While Docsumo excels at financial documents, it also supports general-purpose extraction for logistics documents, healthcare forms, and insurance paperwork. The agent checks your enabled document types and routes each file to the appropriate extraction model based on its classification.