Top 10 RPA use cases with machine learning right now

I am a Chief Product Officer at, a machine learning tool for RPA developers. I view the world through “aito-lenses” so the list contains use cases where our tool can be used, and what I encounter when working with the customers. However, I hope the list serves its purpose for sparking your interest regardless of the tools you use for ML.

2020 was the year of RPA for us at We saw an increasing adoption of our tool as the “brains” of intelligent workflows. We tailored our product to work even better with popular tools like UiPath, Robot Framework and Automation Anywhere. And naturally, we engaged with loads of RPA teams in companies of various sizes and had an opportunity to see — and contribute to — the cool things they are working on.

Here are my top ten favourite use cases where Aito’s predictive power can cut the development time, make maintenance more manageable, and ultimately drive high automation rates that benefit the business.

Let’s dig in! These are not in any particular order.

1. Purchase invoice processing and entry to an accounting system

What: Rule-based automation is a nightmare to maintain. To successfully automate invoices, you’ll need to be posting invoices to the accounting system with the correct details. Use Aito to predict the correct GL account for a new invoice line item, added with cost centre and VAT category (that too often goes missing with OCR).

Why: Faster invoice payments, less repetitive work, fewer errors. Automation can reduce processing times up to 60–80%!

For who: Accounting and finance teams in any larger enterprise.

2. Predict payment times for sales invoices

What: Create automation to predict the “buckets” when each sales invoice will likely to be paid: early, on time, a bit late, a lot late. This allows accounts receivables team to act in a timely matter.

Why: Better prediction accuracy to cashflow planning. Mitigate and act early on likely delayed payments.

For who: Finance and account receivables teams in any enterprise with a big volume of sent invoices.

3. Identify duplicate entries in CRM

What: Various data entry types from system to another is a staple for RPA automation, CRM being one of them. Aito can improve the quality of data at the entry stage, for example, by identifying potential duplicates in CRM entries.

Why: Increased data accuracy, and less manual work fixing the problems.

For who: Sales and marketing and master data teams

4. Categorise and prioritise customer service (or help desk) tickets

What: Common problem for many teams that deal with tickets or requests: how to add metadata such as categories and urgencies to the tickets, so that responses can be better orchestrated? While your RPA or customer service platform takes care of the orchestration, Aito can be the real-time “tagger” that adds labels, categories and urgencies to your tickets based on how your human workers have done in the past.

Why: Faster and more accurate responses to tickets keep customers happy, or at least as satisfied as possible.

For who: Customer service and help desk teams.

5. Amend customer orders with properly categorised service details

What: This is a bit of niche among other cases, but the impact is potentially so massive that it found it’s way to the list. When receiving customer service orders, sometimes paid additional services are not correctly marked at the order time, and details are later amended and manually processed based on delivery team markings. Aito can automate the process by predicting additional service that needs to be added to invoice based on written details.

Why: Increase the revenue by capturing the services that would otherwise go un-invoiced.

For who: Service delivery teams, for example in logistics.

6. Automate product data management in retail/eCommerce

What: Proper tagging of each product is critical for modern retail and eCommerce solutions. Create an automation that looks for new products, missing tags, and uses Aito to predict the likely labels based on product details automatically. Use high confidence predictions without human review, and send the rest for validation. This tireless automation learns from every review made by your staff.

Why: By ensuring the product data is complete, you’ll make sure every product is available for purchase and properly promoted.

For who: Product management teams in eCommerce and retail.

7. Travel and expense claim processing automation

What: This is actually not very different from the #1 on our list, but as I firmly believe every expense claim is time wasted, I chose to include the use case here too. Your workflow can scan the receipts and extract entities, or get details from credit card records. Nevertheless, categories, cost centres and labels (“internal”, “billable” etc.) are needed for claims to be processed and stored, and that is where Aito can help by predicting them based on past entries.

Why: Because everybody hates expense claims.

For who: Accounting/finance and travel management teams.

8. Automate customer invoicing

What: If you have devices deployed to your customers and get usage data (telemetry) that are the basis of invoices, you may be in for some time savings. Instead of manually reviewing the invoices, Aito can predict which invoices are likely to be ok and can be sent automatically, and which ones to hold for manual review.

Why: To save time in mundane tasks and make your employees happier.

For who: Accounting/finance teams.

9. Time record validation in the consulting business

What: More hours are marked to client projects, more revenue consulting business will make. But consultants are generally bad at marking their hours (at least I was!). Automation helps. Create an automation that looks through the marked records in customer projects, uses Aito to predict anomalies. This should alert the account manager and likely culprits to add the missing hours.

Why: To increase the bottom line.

For who: Consulting business leaders.

10. Cross-sell automation of insurance products

What: You’ve got that customer in with the first product, but what then? You are creating elaborate rules of what to sell next. Take a predictive approach! Let Aito predict the most likely customers who do not have a particular insurance product, but have a profile that statistically indicates they should have it. Test sell and feed the results back to Aito for increased accuracy.

Why: Increase customer life cycle value — more revenue.

For who: Sales automation teams

Apart from the benefits listed above, each of the use cases helps you keep your own staff motivated by reducing the repetitive no-brains tasks and making the days more meaningful. That alone is a huge benefit, and so attainable with the tools available today!

Let’s get to work!

Photo by Hal Gatewood on Unsplash




20+ years of SW and tech leadership. Two startup exits. Building next-gen machine learning tools for no-coders and RPA devs at Used to travel a lot

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

My First Project in HTML

I started a Meetup. It was terrifying…

CS373 Fall 2021: Sungwook Kim (Week of 1 Nov— 7 Nov)

Top Issues when Outsourcing Mobile Development

Beginner’s Recommendation Systems with Python

Pipeline Performance Testing with Jenkins and Gatling

Job Roles related to the Database Management

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Tommi Holmgren

Tommi Holmgren

20+ years of SW and tech leadership. Two startup exits. Building next-gen machine learning tools for no-coders and RPA devs at Used to travel a lot

More from Medium

Scraping User Information On GitHub Using PyGithub

Why Is a Full Automation Workflow Necessary for Effective RPA?

Betting on Twitter Sentiment in the NFL, Packers at Lions, Week 18, 2021

Top 5 AI & Machine Learning Frameworks used for Web Development