Title

9 Accounts Payable Problems Solved with AI

Betsy Francoeur

To the surprise of absolutely no one, accounts payable (AP) is ranked as the most laborious, tedious, paper-heavy function in finance — even more so than accounts receivable (AR), tax reporting or payroll.1

Yikes. Thankfully, AP doesn’t have to be such a drag anymore thanks to AI-driven AP automation. Digitizing, automating and applying the latest AI technologies to these essential processes can completely transform this back-office dinosaur into a sleek sports car of a function.

Want to know how? Using Esker’s Synergy AI as an example (that’s what we call our AI technology), here are 9 of the most powerful ways AI is changing AP for the better.

1. Automated supplier onboarding questionnaire creation

An onboarding questionnaire is an important tool for collecting information from your suppliers about their compliance practices, risk management efforts and alignment with your company’s business requirements.

Building such a questionnaire isn’t for newbies. It requires expertise and effort to ensure you’re including all the correct questions for the regions and industries you and your supplier are operating in. Now imagine having to build a brand-new questionnaire every time you onboard a new supplier. Talk about a bottleneck.

That’s where AI comes in. Using a generative pre-trained transformer (GPT) — a type of large language model (LLM) — you can generate onboarding questions tailored to the industry, region and supplier type you need in mere seconds.

How does it work?

Simply prompt the GPT with your general requirements, such as compliance, certifications, risk factors, how many questions it should include, if you’d prefer yes/no or open-ended questions, etc. In moments, the GPT generates a comprehensive, generic due diligence questionnaire that covers all the aspects you provided in your prompt and can be easily edited.

It’s basically ChatGPT for supplier questionnaire creation. Cool, right?

Why should you care?

Automating supplier questionnaire creation is incredibly beneficial for companies with growing supplier networks because it:

  • Saves time and resources: Less manual effort spent on questionnaire development means more time for your team to focus on higher-value tasks like risk assessment and relationship management.
  • Ensures consistency: AI creates standardized questionnaires that cover essential due diligence criteria across all suppliers, reducing variability and maintaining uniformity in supplier data collection.
  • Enhances compliance: Ensures you avoid missing critical compliance-related questions and keeps your aligned with regional and industry-specific regulatory requirements.
  • Scales: AI creates generic questionnaires that can be used across multiple suppliers, simplifying onboarding for suppliers of different sizes and industries.
  • Is user friendly: Anyone on the team can build a questionnaire, even if they don’t have loads of expertise, because AI helps guide the process while ensuring quality.

2. Intelligent data extraction from supplier invoices

Is there anything more tedious than manually keying in invoices? It takes up SO much time, delaying payment processing, diverting resources away from more strategic tasks and filling your team’s days with unending monotony.

Invoice data extraction is one type of work AI really excels in. Using a collection of technologies working together in harmony, it can pull the necessary information off an invoice in a flash.

How it works

Esker uses four layers of technology to extract data from invoices: character extraction, first-time recognition, auto learning and teaching.

  • Character extraction pulls out relevant information from the invoice using intelligent data capture technology like optical character recognition (OCR) and deep learning capabilities.
  • First-time recognition extracts data from invoices seen for the first time by identifying predefined fields on the document and, if configured, comparing that data against historical documents and ERP master data.
  • Auto-learning is a machine learning (ML) technology that automatically learns from user corrections to increase future recognition rates.
  • Teaching lets you explicitly define recognition criteria for specific recurring invoices so that they can be accurately recognized every time.

This 3-page overview explains intelligent data extraction in more detail.

Why should you care?

What’s not to love about technology that takes over the soul-sucking grunt work of manual data entry from supplier invoices?

  • Increased speed and efficiency: AI can process large volumes of invoices in seconds, significantly reducing turnaround times and ensuring timely payments.
  • Accuracy and consistency: AI can extract data accurately, even from complex layouts and unstructured documents — plus machine learning helps it keep improving over time.
  • Cost savings: Automated data entry reduces labor costs and minimizes the impact of errors or delays.
  • Scalability: Handle a growing workload without adding additional headcount.
  • Improved employee satisfaction: Your team can focus on higher-value tasks, improving job satisfaction and reducing burnout from repetitive, low-value work.
  • Real-time insights: Always have accurate, up-to-date financial information for smarter decision-making.
  • Enhanced compliance and tracking: Integrated workflows and detailed audit trails ensure proper processes are followed and all necessary approvals are received.

3. Automated invoice batch separation

Scanned or emailed invoices often come in batches of multiple invoices in one big file. Manually separating batches into individual invoices is yet another menial, time-consuming task that AI can take over.

How it works

Using deep learning technology, Esker’s AI trains itself to automatically recognize individual invoices within batches and suggest splits that a user can review and confirm. The AI can do this even if some invoices are multiple pages long or have attached documents, keeping all related pages and documents together during the splitting process.

Why should you care?

The biggest benefit of this use case is saved time. When your team doesn’t have to painstakingly split batches of invoices, those invoices can be processed and paid faster. Plus, staff can spend their time on bigger picture tasks that bring more value to your business.

4. Predictive supplier invoice coding

Non-PO invoices are the bane of every AP department. But they’re a fact of life, especially in spend categories such as office supplies or admin costs that often require splitting the invoice cost among different cost centers and G/L accounts.

Sorting through this mess manually is not only time-consuming, but the potential for errors is high. AI-based predictive coding can suggest the correct coding for non-PO invoices for you. (Fun fact: Esker actually has a U.S. patent on this innovation.)

How it works

The AI observes and learns from a user correctly coding non-PO invoices. From there, it uses those examples
to recommend applicable coding for future non-PO invoices. The AI continues to learn from user corrections and continuously improves its prediction accuracy. Typically, after only a few repetitions, non-PO invoices are automatically allocated to the correct G/L account, cost center, cost type and tax rate.

Watch it in action in this quick video.

Why should you care?

Accuracy is the key benefit here, not to mention efficiency. Faster invoice processing leads to more on-time payments and happier suppliers.

5. Anomaly detection in supplier invoices

When an uncommonly large or small amount shows up on a supplier invoice, it’s important to flag it so you can confirm if it’s legitimate or a mistake (we’ve all fat-fingered something on our keyboards — no shame). AI can spot those in flash. (Bonus: it can spot anomalies in purchase requisitions as well).

How it works

You can set up Esker’s Synergy AI to automatically warn you in case of unusual quantities or amounts for a given item and/or supplier. It will analyze your historical order data and invoices to detect anomalies in payment amount or order quantity and alert you.

Why you should care?

Instead of your staff having to dedicate their valuable time to poring over purchase requisitions and supplier invoices hunting for oddities, AI takes on the task — and does it faster and more accurately. Identifying all anomalies decreases your risk of placing an incorrect order or making an erroneous payment.

6. Late payment predictions

When you’re managing hundreds, if not thousands, of supplier invoices every month, you run the risk of missing payment due dates. Nobody wants to be THAT customer who gets a reputation for being a late payer. Plus, not paying your invoices on time can lead to blocked shipments and other unpleasant consequences. Thankfully, AI can help you keep an eye on due dates so you never miss another payment.

How it works

The machine learning within Esker’s Synergy AI uses historical data (such as the supplier, the type of items and users involved in the workflow) to predict the likelihood of an invoice being paid late. It creates a list of these at-risk invoices that displays on the user’s home screen. The user can then drill down to the actual invoices to see what’s holding up the process and take action to get them moving.

Check it out in this 50-second video.

Why should you care?

Not paying your invoices late is a great way to get more early payment discounts, avoid late payment fees, reduce the risk of shipping delays and maintain good supplier relationships.

7. AI-assisted email triaging

Even if you have one dedicated email inbox for all supplier communication, sorting through it to determine which emails are questions, invoices, status update requests or payment reminders can feel like a full-time job by itself. Why not let AI help you out?

As part of Esker’s Supplier Inquiry Management solution, AI can triage incoming emails to separate out the invoices (which we’ll explain in this section) AND generate appropriate answers to any incoming emails (explained in the next section).

How it works

Using AI and natural language processing (NLP), Esker automatically identifies any invoices received in your AP mailbox with 95% accuracy. Non-invoice emails (a.k.a. inquiries) are also identified and classified. Emails are seamlessly and automatically redirected to the Esker platform for processing or response.

Why you should care

AI-assisted email triaging significantly enhances the efficiency and accuracy of AP teams by:

  • Speeding up invoice identification and processing
  • Providing transparency with trend analysis and performance metrics
  • Increasing employee satisfaction by eliminating manual email triage
  • Improving vendor relationships with quick responses and on-time payments

8. Generative AI-assisted responses

When your suppliers email you, they expect a quick response that thoroughly answers all of their questions. But that’s easier said than done. Good responses to supplier inquiries means sorting out the requests so they can be routed to the right people, researching internal systems for needed information and drafting helpful replies. It’s labor-intensive, time-consuming work that often results in frustrated employees and annoyed suppliers.

As mentioned in the previous section, Esker Supplier Inquiry Management not only triages incoming emails so requests are separated out from other types of emails, but it can also find answers and draft responses for you.

How it works

First, the AI reads the email to determine what’s being asked. Then, if necessary, it searches your company’s internal data for the answers. Finally, it drafts a response using a ChatGPT-type technology. Using this AI-enabled tech means the responses aren’t static, template-based replies, but dynamic, context-aware answers tailored to the specific needs and emotional tone of each request.

An AP team member can review the AI-generated response to ensure accuracy and relevance. You can either send it as-is or personalize it even further to add a human touch — either way, you’re always in control. You can see it action in this video.

Why should you care?

Leveraging AI to respond to supplier queries:

  • Accelerates response times so your suppliers aren’t left hanging
  • Delivers consistent and accurate answers that build trust and satisfaction with your suppliers
  • Enhances team productivity and morale by reducing time-consuming, repetitive tasks

9. Esker Synergy Digital Assistant for AP staff

Even if all of your sourcing, procurement and accounts payable information is digitized, locating the exact information you need in a pinch can still feel overwhelming — or even worse, like finding a needle in a haystack. Whether you’re trying to create a custom report or need to know which invoices are in a specific status, AI can stop the stress by helping you pull together the data you’re looking for.

Meet Esker’s Synergy Digital Assistant. Think of it as a know-it-all coworker eager to help you find and organize information.

How it works

Open the chat window and type in your question or request. You can even use voice-to-text to ask your question if you prefer. Then Synergy Digital Assistant goes to work. In seconds, this AI-powered helper processes your request and responds within the chat. Answers are typically in the form of a graph, a filtered list or information pulled from documentation.

Say you want to find out which invoices are in set-aside status. Type in your question and the assistant pulls up a list of invoices in that status. From there, you could ask it to sort the list in ascending or descending order. You could also ask it to tell you the average DPO of these set-aside invoices. You can even download any graphs or reports the assistant generates for you.

And that’s just the tip of the iceberg. Check out this short video to see more of what the Synergy Digital Assistant can do.

Why should you care?

Besides the fact that having your very own assistant just sounds neat, an AI-powered assistant gives you near-instant access to useful analytics, metrics and documentation about your automation solution. It helps newer AP specialists get up to speed faster because they can simply turn to the assistant when they have questions. And on top of that, it’s a new and enjoyable way to interface with your AI-driven automation tool.

Bring the power of AI to your AP department with Esker

Manual, paper-based AP processes are so last decade. Finance teams who want to stay competitive and agile in a shifting world are embracing AI-driven solutions today so they don’t have to play catch-up later on.

We can help. At Esker, we do AI differently.

Based on more than 15 years of research and development, our Synergy AI combines GenAI, machine learning, OCR, robotic process automation (RPA), intelligent data processing (IDP) and other technologies that all work together seamlessly to deliver a truly automated experience that dramatically reduces or even eliminates the need for manual intervention at any point in the AP process.

The GenAI in Esker is securely hosted to protect your valuable data and ensure it’s never accessed by third parties. And while AI is important, at Esker, we only use AI in targeted, purposeful and value-added scenarios to reduce its ethical and environmental impacts.

While many solutions exist that can automate piecemeal parts of the AP process, we’ve created AI-based solutions for the entire source-to-pay and order-to-cash cycles. So you can bring the AI advantage to all facets of your financial operations.

See for yourself by visiting our demo library to learn what AI-enabled AP and AR processes look like.

Sources
1 2023 Benchmarking Report: Automation Trends, IOFM

Author Bio

Betsy Francoeur

As a Copywriter at Esker, Betsy loves writing about the source-to-pay and order-to-cash cycles and creating valuable content for financial professionals. She also enjoys running 5ks, kayaking, traveling with her husband and snuggling her dog.

Undefined
Author Photo: 
Top