The manual process of matching invoices to purchase orders (POs) involves several steps and can be time-consuming, especially for companies that receive multiple invoices and POs daily. Here's a breakdown of the manual process:

1. The manufacturing company creates a purchase order (PO) when they want to purchase raw materials from a supplier. The PO includes details such as the product name, quantity, unit price, discounts, shipping, and the total amount.
2. The supplier delivers the goods and then sends an invoice to the manufacturing company.
3. When the manufacturing company receives the invoice, an employee searches for the corresponding PO number (in this case, 20231218).
4. The employee then compares the details on the invoice (unit price, quantity, discounts, shipping, and total) to the details on the PO to ensure that everything matches.
5. If all the details match, the employee approves the invoice for payment. If there are any discrepancies, the employee must investigate and resolve the issue before approving the payment.

This manual process can be tedious and prone to errors, especially for companies that handle a large volume of invoices and POs. Automating this process can help streamline the workflow and reduce the risk of overpayment or underpayment.

The process of manually matching invoices to purchase orders (POs) can be time-consuming and error-prone, especially for companies that receive multiple invoices and POs daily. Vector Shift provides a solution to automate this process, ensuring accurate matching and streamlining the payment process.

The key steps in the automated process are:

1. Extract Invoice Data: The pipeline extracts the relevant data from the invoice, including the PO number, vendor name, total amount, and a list of all the items purchased.

2. Match Items to PO: For each item in the invoice, the pipeline matches the quantity and unit price to the corresponding PO data. This is done by querying the PO data stored in a knowledge base and comparing the values.

3. Validate Total Amount: After verifying the individual item matches, the pipeline checks if the total amount on the invoice matches the total amount on the PO.

4. Output Matching Result: The pipeline outputs a message indicating whether the invoice matches the PO completely, or if there are any discrepancies in the item quantities, unit prices, or total amount.

By automating this process, companies can save time, reduce errors, and ensure that they are paying the correct amounts to their suppliers. The Vector Shift platform provides a flexible and scalable solution to handle this advanced use case, allowing companies to focus on their core business activities.

To automate the process of matching POS to invoices, we first need to extract the relevant data from the invoice. The key data points we need to extract are:

1. PO Number: This will be used to find the corresponding purchase order in the knowledge base.
2. Vendor Name: This provides additional context about the invoice.
3. Total Amount: The total amount on the invoice needs to be matched against the purchase order.
4. List of Items: We need to extract a list of all the items purchased, as we will need to match each item individually against the purchase order.

For each item in the list, we also need to extract the following:

1. Quantity: The quantity of the item purchased.
2. Unit Price: The unit price of the item.

By extracting this data from the invoice, we can then systematically match each element against the corresponding purchase order to ensure the invoice is accurate and can be approved for payment.

When a manufacturing company purchases raw materials, the process typically involves the following steps:

1. The manufacturing company creates a purchase order (PO) and sends it to the raw material supplier.
2. The raw material supplier delivers the goods and sends an invoice to the manufacturing company.
3. The manufacturing company must match the invoice to the corresponding PO to ensure the correct quantities, unit prices, discounts, and total amount are billed.

This manual process can be time-consuming and error-prone, especially for companies that receive multiple invoices and POs daily.

To automate this process, the company can use Vector Shift to:

1. Extract the relevant data from the invoice, including the PO number, vendor name, total amount, and a list of purchased items.
2. For each item in the invoice, determine the quantity and unit price.
3. Match the invoice item details (item name, quantity, unit price) against the corresponding PO data.
4. Verify that the total amount on the invoice matches the PO.

By automating these steps, the company can streamline the invoice-to-PO matching process, reducing the risk of overpayment or underpayment and freeing up employees to focus on other tasks.

To validate the total amount, the pipeline performs the following steps:

1. It retrieves the PO data from the knowledge base using the PO number extracted from the invoice.
2. It compares the total amount in the invoice to the total amount in the PO data.
3. If the total amounts match, it outputs "yes".
4. If the total amounts do not match, it outputs "no".

The pipeline uses an OpenAI node with a prompt that instructs it to match the total amount in the invoice to the PO. If they match, the node outputs "yes", otherwise it outputs "no". This output is then used in a conditional node to determine the final output of the pipeline.

If the list of items contains at least one "no" response, indicating that the item quantity or unit price does not match the purchase order, the pipeline will take the "no" path in the condition node. In this case, the pipeline will output the message "item quantity and/or item unit price does not match".

If the total amount in the invoice does not match the purchase order, the pipeline will take the "no" path in the second condition node. In this case, the pipeline will output the message "total amount does not match".

If all items match and the total amount matches, the pipeline will take the "yes" path in both condition nodes and output the message "total amount, item quantities, and unit prices all match".

The pipeline uses a merge node to combine these three possible output paths and a final output node to display the appropriate message based on the results of the matching process.

The process of matching POS to invoices can be a time-consuming and error-prone task for enterprises. However, by leveraging the capabilities of the Vector Shift platform, this process can be automated and streamlined, leading to significant time and cost savings.

The key steps involved in this automation include:

1. Extracting relevant data from the invoice, such as the PO number, vendor name, total amount, and list of items.
2. Matching each item in the invoice to the corresponding PO, ensuring that the quantity and unit price align.
3. Verifying that the total amount on the invoice matches the PO.
4. Providing clear feedback on whether the invoice matches the PO, or if there are any discrepancies.

By automating this process, enterprises can reduce the manual effort required, minimize the risk of errors, and ensure that payments are made accurately and on time. This not only improves the efficiency of the accounts payable process but also enhances the overall financial management of the organization.

The demonstration provided in this video showcases the flexibility and power of the Vector Shift platform in addressing complex business challenges. By breaking down the manual process into a series of logical steps and automating them using the platform's features, the enterprise was able to achieve a seamless and reliable invoice-to-PO matching solution.