AI for executive assistants: Lessons from Lori Propp-Anderson, EA at GitHub

At GitHub, manager and executive assistant Lori Propp-Anderson has built a career around supporting leaders with foresight, empathy, and adaptability. Today, she’s also leaning heavily into how technology, especially AI for executive assistants, is changing the role.

This blog is part of our series spotlighting top EAs in tech, and we were fortunate to sit down with Lori to hear how she uses AI for executive assistants to work faster, smarter, and more effectively.

Table of contents

1. How do you currently use AI to maximize your productivity as an EA?
2. What’s something AI has made dramatically easier for you, and what’s something it still can’t replace?
3. If you could design the perfect EA toolkit five years from now, no limits, what would it include, and what would you happily never do again?
4. What can executive assistants do that AI will never be able to replicate?
5. What’s the secret to building a successful EA/executive relationship?
6. What’s the most difficult situation you’ve had to handle as an EA, and how did you approach it?
7. What’s one decision or change you made in the background that your executive might not even realize completely changed the game for them or the company?
8. What advice would you give to an executive considering working with a remote EA for the first time?
9. What’s one thing you wish you had known in your early days as an EA?

1. How do you currently use AI to maximize your productivity as an EA?

Oh, honestly, in so many ways. There are about three AI tools that I rely on. Within GitHub and Microsoft, I use Microsoft 365 Copilot, GitHub Copilot, and Gemini.

I provide communications support to my executive, and with AI using the right prompts, I can quickly create a draft to get me started on what I want to write for him. I’ve even asked AI to write in the style of my executive because he has enough public content and internal content for AI to really understand how he writes.

When you’re working in a company like GitHub, you have to think globally. I lean on AI to adjust writing for different cultural norms. For example, if I’m preparing a message for an audience in Japan, business communication looks very different from what it does in the United States. AI helps me shape the message so it resonates appropriately with that audience.

Another way I’ve been using AI is creating a lot of process documents as I build a portfolio for succession by turning demos into SOPs. One of the hacks that I’ve been using is I just create a quick Zoom meeting, record a demo of what I’m doing, and then put the transcript into AI and ask it to create a process document in the format I want. It delivers it right up without me having to spend the time typing. From there, I can polish, add links, and finish things up for publication.

AI has multiplied my output in creating process documents for the admin community. One thing I want to have completed within the next year is an AI agent for the EA community that can reference all this content, so they can just ask it a question and it will serve up the answer.

Can you share a creative way you’ve used AI in your role?

I’ve used AI to generate the code I need to create interactive Slack messages with buttons and links, something I never could have done on my own.

There’s a tool that works with Slack where you can create messages with linkable buttons. It’s done with code, but I’m not a coder. I can take the message that I want, put it into M365 Copilot, and say, ‘I want this in JSON code with emojis and a button that links to content.’ It writes the code for me. I just copy it into Slack Block Builder Kit and send it.

Then I created a process document on how to do it and shared it with all the EAs in my org so everybody can do it easily. Some AI plays are for quantity, some are for quality. This one takes the same amount of time, but the output is more engaging for the audience.

*This is what code turns into when you use Slack’s block kit builder.

2. What has AI made dramatically easier for you? And what can it still not replace?

AI has made writing dramatically easier for me. At our revenue kickoff event, I had to write 23 award scripts. I gave AI the parameters: anonymized, 700 characters or less, consistent style, etc, and uploaded the nominations. It instantly wrote me scripts that would have taken eight hours. I was done in less than one.

What can’t it replace? Emotional intelligence. AI can’t manage relationships or be that bridge between you and your leader, their directs, the team, customers, and peers. It can’t read the room or bring empathy. That’s our bread and butter. I’m also the culture champion for our org, and while I lean on AI for ideas, we’re the ones that bring culture to our teams.

3. If you could design the perfect EA toolkit five years from now, what would it include?

I would love to see multiple agentic AIs: automated helpers that can do specific tasks. For example, planning an event with sourcing, menus, and accommodations. Or an agent that can build a scope of work and create purchase orders.

The dream? An agent for expense reports. Upload the receipts, and it creates the report. That’s something I could seriously get behind: never having to do an expense report again.

4. What will executive assistants always do better than AI?

AI is transforming the workplace, but there are aspects of the EA role that remain distinctly human. I’m not just managing projects and logistics. I’m managing relationships and trust. I anticipate needs before they’re spoken. I read the room when there are tensions. I adjust communication styles to personalities and priorities.

AI can suggest a meeting time, but it can’t sense when an executive needs space or when you need to hop on Zoom with another EA to solve scheduling conflicts. It can’t do that.

5. What is the secret to building a successful EA–executive relationship?

It’s about building relationships versus workflows. The best partnerships are built on mutual respect and rapport.

One of the first things I do with a new leader is ask if they’ve ever done a personality profile like DISC, Insights, or Myers-Briggs. I want to know their working style and share mine. That way, we both understand how we communicate and complement each other. It’s helped me build strong relationships over the years.

6. What’s the most difficult situation you’ve had to handle as an EA, and how did you approach it?

I have a somewhat recent example. At our revenue kickoff, like most big events, there were a lot of cooks in the kitchen. As an EA, I wasn’t just the point person: I was the project manager and ultimately responsible for the outcome of a section of the event.

Based on past years, I knew the schedule was likely to slip. So this year I didn’t just have a plan B, I had a plan C and D. I mapped out all the different ways the presentation could go off track. I worked with the planning team and presenters so everyone knew the backup plans and the signals to pivot.

We did pivot on the fly, like clockwork. What could have been a chaotic scramble turned into a seamless shift. The presenter stayed calm, the team stayed aligned, and the audience never noticed. That’s the magic of proactive planning.

7. What’s one change you made that your executive might not even realize completely changed the game for them or the company?

While analyzing our budgets, I found hundreds of thousands of dollars that had been inadvertently charged to our Travel & Expense and morale budget. They were valid charges but in the wrong GL account code. I reclassified them, which allowed us to still make the right travel decisions because the budget was available.

I’ve also discovered people in the wrong cost centers, which meant payroll and expenses were hitting the wrong groups. That’s the kind of impact EAs can have. We manage the details, spot the patterns, and fix the friction to make sure the machine runs smoothly.

8. What advice would you give an executive working with a remote EA for the first time?

It’s been nearly 15 years since I supported someone local to me. My top advice is to keep communication open and not just by email. Video matters. When you think you’ve communicated enough, communicate more.

At the start, expect to spend time building trust. Define expectations and outline what success looks like. Almost like an AI prompt: the more clarity at the start, the more on target the output will be.

Empower ownership. At GitHub, autonomy is valued, and it unlocks impact. Keep us in the loop so we can anticipate and represent our leader the best.

9. What do you wish you had known in your early days as an EA?

I started in 1987. I wish I had known about email, smartphones, even the internet. But honestly, what I wish I had known earlier is the value of personality profiles and cultural training.

At the end of the day, it’s about people. Understanding interpersonal dynamics and cultural aspects transcends technology and time.

Lori shows how experience and innovation can coexist. Her work proves that AI for executive assistants isn’t about replacement, it’s about extending impact, creating scalable processes, and unlocking time for the human skills that make EAs indispensable.

Turn AI into an advantage with the right support. Book a call today to get an EA experienced in applying AI for executive assistants.

Fadua Caicedo Díaz del Castillo

Fadua is a bilingual advertiser and holds a master’s degree in creative writing. With over ten years of experience, she has written countless advertising and social media campaigns, blogs, interviews, and everything in between. She writes about startups, the impact of executive assistants, and the stories behind their work. When she’s not writing, she is spending quality time with her husband and son, hiking, reading, or discovering new cafés.