3 AI Case Studies for Equity Plan Automation

Perhaps you’ve been tasked with finding ways to use AI to streamline management of your company’s equity program, but you aren’t sure where to start. Previously, I’ve discussed general ways to use AI for equity plan management (see my blog “Could a Robot Manage Your Equity Plan?” and my video “3 Ways AI Can Streamline Equity Plan Management”).

In this blog, I get more specific with three real-world examples of AI in stock plan management, presented during the NASPP AI Week program, AI for Equity Compensation. If you are looking for opportunities to automate stock plan processes with AI, these are three great places to start.

Complete Tax Reconciliations for Equity Plan Transactions

Cindy Croom of Walmart uses AI to automate RSU vesting reconciliation, validating the tax rates used for each vesting event against a file provided by Walmart’s payroll team. Before AI, this was a labor-intensive, manual process involving copying and pasting records into multiple Excel worksheets and using formulas and pivot tables to cross-reference the data.

Cindy used AI not only to automate these processes but also to build a dashboard that allows her to initiate the process with the click of a button, almost immediately see the results of the reconciliation, and drill down into any errors. Here are some things Cindy can do using this dashboard:

• Compare the report of vesting events from Walmart’s third-party administrator to the tax rates provided by Walmart’s payroll team to identify discrepancies.
• Compare the results of this reconciliation with a report of participant tax elections, which may resolve some of the discrepancies.
• Identify transactions by mobile employees that need to be forwarded to Walmart’s third-party advisor that assists with tax compliance for these employees.
• Create an upload file to override the fair market value (FMV) for specific types of transactions, such as vesting events for employees in India. The file is created in the exact format necessary to submit to Walmart’s third-party administrator.
• Create and download the spreadsheet the team used to create manually.

It took Cindy just under 20 hours to create this AI-powered reconciliation process and dashboard, but she estimates that it saves Walmart’s equity administration team around 180 hours annually. She describes how she created the process and dashboard in the session “Stop the VLOOKUP Madness: Automating Vest-Day Validation with AI” during NASPP’s AI for Equity Compensation program.

Create Journal Entries for Equity Plan Transactions

Journal entries, which are used to record financial transactions, such as equity plan transactions, in a company’s general ledger, can be tedious to create. Connie Zeng of Microsoft uses AI to automate journal entries with a Visual Basic for Applications (VBA) script she built for her team’s equity plan transactions. Just like Cindy’s reconciliation process, before AI, creating these journal entries was a manual process: data was copied and pasted from several different Excel files, then formulas and pivot tables were used to calculate the entries and manipulate them into the appropriate format.

The script Connie was able to create using AI does all this work. Connie downloads a master report from Microsoft’s stock plan system, then runs the VBA script to complete the following:

• Supplement the report with the information necessary for the journal entries.
• Create the journal entries.
• Output the entries in a format that is ready for upload to the system that records Microsoft’s journal entries.

Connie demonstrates how she used AI to create the VBA script in the session “Advanced Spreadsheets to Agentic AI: Intelligent Automation Opportunities in Equity Operations” during NASPP’s AI for Equity Compensation program. Her slides include some great examples of the prompts she used to direct Copilot to create the script and fix issues with it.

Collect Travel Data from Mobile Employees

Like any company that offers equity in China, Amazon has to comply with State Administration of Foreign Exchange (SAFE) requirements. For Joleen Lawson, Amazon’s global stock plan specialist, this includes collecting travel data from non-Chinese nationals in China. Previously, Joleen managed this process through email. Because all the data was collected at once, this was cumbersome, and it would take her several days to parse through it all.

To automate the collection process, she used an internal AI tool to create a Slack bot (i.e., a bot in the collaboration tool Slack) that would reach out to employees in China and collect the data from them. Here is what the bot does:

• Proactively reaches out to employees to collect their travel data.
• Asks employees a series of questions designed to ensure they are providing the correct data. Employees can also choose to download and complete the forms on their own and use the Slack bot to help them verify that they’ve completed the forms correctly.
• Performs real-time validation on the data employees provide. For example, the bot excludes travel that is outside the appropriate time period.
• Provides immediate feedback to employees about whether they will be eligible to register for fund repatriation.

Joleen says that this has shortened the task from three to five days of manual review to a few minutes of employee interaction with the bot. She also notes that it will be much more scalable as Amazon’s population of employees in China grows. Joleen demonstrates how the bot works in the session “How Equity Teams Are Applying AI in the Real World” during NASPP’s AI for Equity Compensation program and includes the prompts she used to create the bot in the supplemental materials for the session.

Tips for Using AI to Create Successful Equity Plan Processes

Below are a few helpful tips for interacting with AI tools that I learned while listening to these case studies.

Describe What You Want in Plain English

You don’t need to be a coder to use these tools. You simply explain what you want in plain English, and the AI tool writes the code for you. It’s helpful if you give the tool as much context as possible. Explain what you are doing and why. If you hit a roadblock, for example, if the code doesn’t work or you can’t figure out how to deploy it, ask the AI tool for help.

Be Careful with Personally Identifiable Information (PII)

If you will need to upload PII to the AI tool you are working with, make sure you are using an enterprise version of the tool and that this is permissible under your company’s policies. The good news is that you do not need to upload PII to use these tools. For example, Claude can write a VBA script that you deploy in Excel without ever seeing any participant PII. You simply tell Claude what you need the script to do, and then copy the script it creates into your Excel file.

Use AI to Brainstorm Solutions

Sometimes you aren’t sure how AI can help streamline a process. You can present the problem to AI and ask it to suggest solutions. Provide as much context as possible and include any restrictions (such as a limitation on uploading PII). If the solution it suggests doesn’t seem feasible, explain why and ask for a different solution. You can even ask AI to suggest appropriate AI tools to use to build the solution.

Run Processes in Parallel to Evaluate the AI Solution

To evaluate whether your AI process is working correctly, continue to use your manual process and compare the results to the AI process results. Cindy notes that Walmart’s team did this for several vesting cycles until they were comfortable with the AI process. This not only helps you and your team feel confident about the AI process, but it can also help demonstrate to your auditors that the process is effective.

Learn More with NASPP’s AI Week Program

These three case studies are just a small sample of the ideas presented during our AI for Equity Compensation program (now available for on-demand viewing), covering AI tools for stock plan administrators. I counted at least 15 total case studies, many of which include demonstrations of how the processes were created and samples of the prompts used. Check it out today!