Automating Accounts Payable with Blue Copa

Blue Copa utilizes artificial intelligence to streamline and automate both PO-based and non-PO-based accounts payable processes. This document provides a step-by-step walkthrough of the process, highlighting how the AI-driven platform manages tasks from data ingestion to reconciliation.

Step 1

Begin by navigating to the Account Payable dashboard within the PO-based context of Blue Copa.

Step 2

The platform consists of three main pages: the preparer module, the approver module, and the reconciliation result page. Start by using the preparer module to handle data ingestion.

Step 3

In Blue Copa, data can be ingested through three methods: an AI-powered email agent that scans emails for relevant files such as POs or invoices, native connectors that integrate with your ERPs to transfer data seamlessly, and a manual upload option for files like POs and invoices.

Step 4

Once files are uploaded, Blue Copa's AI-driven vision model scans the documents to extract relevant information and creates a reusable dataset. This model standardizes invoices across different languages and formats. In the preparer view, a summary of ingested invoices is provided, indicating valid and invalid invoices.

Step 5

The vision model assesses the validity and authenticity of invoices based on parameters such as amount, threshold errors, and line item inconsistencies. Valid invoices are then forwarded to the approver for review and approval.

Step 6

The agent categorizes rejected invoices into different buckets, such as duplicates or missing vendor information. Despite these categories, the approver has the option to approve or accept invoices manually. Once approved, the data is fed into Blue Copa's AI-driven reconciliation module.

Step 7

The AI identifies columns that require reconciliation in both the PO and invoice datasets. It aligns relevant columns from the PO with those on the invoice and executes the reconciliation process. Results are displayed on this screen.

Step 8

An agent operates in the background, providing insights into processed invoices and records. It details the matched and unmatched records, the next scheduled reconciliation, and the last reconciliation date. Scrolling down reveals both matched and mismatched invoices.

Step 9

By scrolling to the right, executives can take corrective actions on mismatched invoices.

Step 10

Executives can view data analyzed by the Blue Copa vision model, featuring a side-by-side comparison of PO data and invoice data. Even for mismatched entries, the approver can accept, approve, or reject the invoice and submit their decision.

Step 11

This concludes the process for PO-based accounts payable.

Step 12

Next, let's explore the process for non-PO-based accounts payable.

Step 13

In the invoice action agent screen for non-PO invoices, users can drop invoices directly. The AI email agent can automatically identify relevant files from emails and feed them into Blue Copa.

Step 14

Alternatively, data can be ingested via connectors or manually. Once invoices are dropped into Blue Copa, a task is created for the preparer to review and take necessary actions.

Step 15

The table is updated with the latest invoices. In the approver view screen, approvers can take action on invoices using the task table.

Step 16

After actions are completed, invoice data can be sent to the treasury team with a single click, enabling them to commence processing. This exemplifies how Blue Copa uses AI to create an efficient accounts payable flow. Thank you for your attention.