Andrea Silas- Two Heads Are Better Than One: How AI Can Work Together with Human Agents to Provide Superior Customer Support
Andrea Silas, the Vice President of Technical Support at DreamHost, is a seasoned professional with nearly 20 years of experience in the company. Starting as a Technical Support team member, Andrea has been promoted four times and now leads the team as the VP. She has played a key role in scaling and reorganizing the support team, resulting in a high customer satisfaction rating.
Watch this conversation video featuring Andrea Silas our upcoming speaker for November's Leadership Summit, where She delves into the exciting topic of utilizing AI to enhance your technical support team and elevate customer assistance. This video offers a sneak peek into the compelling insights Andrea will be sharing on stage in Oakland. Be sure to watch and get a taste of what's to come!
Andrea Silas:
I am Andrea Silas. I run the tech support team at DreamHost. We are a web hosting company that's been around forever.
This is my 20th year in support, and I have been the vice president of the technical support team since 2014.
Scott Tran:
Let's talk about AI. It seems like a lot of companies are taking a, throw it against the wall and see what sticks approach. So, I was curious about how are you thinking about it. How does it fit into what you're doing?
Andrea Silas:
After the initial, what is happening moment? And then trying to catch up with my boss, my CEO, who is like super technical. He's a developer, so he is always ahead of the game and he is like, we are gonna be doing this now.
Let's start thinking about AI. You should be using AI for everything. Whoa, okay, let's figure out how we can actually use AI that benefits the team. The first idea was basically getting ahead, what looks like a threat to what we do, in support because AI is everywhere.
The news covers AI from all perspectives, and then you start seeing things like companies are not hiring because of AI or they're laying people off from support from ai and that's not what we want to do. So, let's figure out how we can take advantage of AI for the benefit of the team.
How can we make the team better? How can we use AI to make the people of the team better rather than substitute the team? What we decided to do, is augment the team with AI. I know there are a lot of companies who are like, how we can do this customer facing and that's not what we wanna do or we don't think it's there yet from our perspective.
So, what we decided is to try to figure out with all the things that we want our team members to do while they are assisting customers, how can we make use of AI to help the process.
One of the first things that I did when I realized that this was happening and we are actually doing something about it, is to communicate very clearly and openly with my team to explain what is happening and how we are going about it, and how they are going to be involved.
So as a leader, you wanna make sure that your team knows what's going on, be it good or bad.
And you need to communicate because this is, going to affect the teams, in one way or another.
Scott Tran:
You mentioned that your CEO is very technical and how did that conversation go how should we as a company be using AI?
Andrea Silas:
He is pushing us always toward solutions. We're doing really great, but how can we do even better? What can you do with the team? And what are the tools that we can use? If something is out there, how can we take advantage of it? And because he is very technical, the company itself is very technical.
We have a lot of developers, a lot of curious, people who basically took it up on themselves like, we're gonna help support.
So as soon as we heard from the CEO that we are doing this, I got a meeting with, one of our developers who was super interested in putting time aside from all the projects that we are gonna build something for support, and we're gonna call it CS Assist, which is basically indicating that this tool is going to get created for tech support with the validation, input from tech support or what we need, how can we use it, how we would benefit from it, and we are gonna build it internally. That's what we've been doing.
Scott Tran:
How did you get up to speed on what's out there, what's possible what kind of things would you like to have?
Andrea Silas:
First, like really understanding what is possible and isn't possible, what AI actually means from the ChatGPT perspective, having conversations with my internal dev made that possible.
And then once I understood what the limitations and the opportunities are, then we started compiling ideas of what would be quick wins.
To get started and to get the buy-in from the team. And then once we had the list of what is possible and what would be quick to do and what would show, a quick win, a quick result. Once we had that, we started working on it. And in the meantime, we are already creating phase two, phase three to go all the way to everything that we want.
And all the people that have been pulled in from all levels of tech support are super enthusiastic about it.
It's, I want this, I want this to do this. This would help me. This is how I would be able to help the customer faster, or not have to do the groundwork, the prep work, and something else can do it for me.
Then I can focus and actually help the customer rather than reading pages of support, history and, searching the knowledge base.
Scott Tran:
What are some of the things that you're seeing, is this possible with AI? Essentially what are the strengths and weaknesses of AI?
Andrea Silas:
So, we already ran into some limitations. For example, one of the things we want AI to do is recognize the workflow that is required to do a certain level of assistance. Our workflows right now are all over the place. They're documented here and there, or not documented at all. AI is not a magician. It's not going to recognize all the sources and then create a workflow for us. What it can do is use a workflow once it matched the topic. So what we have to do, we have to do our own legwork ahead of time.
So what we are doing right now is creating the workflows in a certain format. Then AI can use that and pass it to the person who requires that workflow while they are trying to assist. So it's not like a human who can just read through something and figure out what it is and then do the answer, we have to do the prep work for the AI to be useful.
Scott Tran:
Was it challenging to get developers to help you build these things?
Andrea Silas:
No, they were ahead of me. So by the time I got to the point where, all right, this is happening, somebody was already, compiling ideas from their perspective and it's really great that you wanna help. So let's make sure that those two things align and then we had to do like a sanity check and a validation of the stuff that they came up with. What would be beneficial? And we're like, yeah, those are cool, but I, we feel like these, and this, and this would be a good thing to attack first.
We have to accommodate the fact that these are the same developers who work on our products and infrastructure and whatnot. So, there is a little bit of delay, but because we are building it internally and everybody is super enthusiastic too to make it happen. We're still on time.
Scott Tran:
What's exciting for you about adding AI to your tool set or having it be a part of how you work?
Andrea Silas:
Doing support can be a little bit of drudgery sometimes when you have to keep answering the same kind of issues over and over before you can answer the question appropriately.
We have a lot of expectations for the team and that means, reading all the support history, and being aware of the 15 different kinds of maintenances that are happening right now. Or, being aware of the workflows and the limitations of those workflows, and all those things can be caught out.
And something can deliver that for you. So all you need to do is use your knowledge, use your expertise, use your human ingenuity to help that customer the way that customer needs to be helped, and do it quickly. So they can now focus on becoming as technically helpful for the customer as they can without having to be bogged down, so to speak.
It's going to increase team happiness. It's going to reduce burnout. It's going to open opportunities for team members to find things that they are interested in and then allow time for them to discover those interests.
It's going to increase customer satisfaction because, getting your answer, quickly and effectively, the way that it's supposed to happen, and the way you expect the company to assist you, that is very important for the customers.
It's for the customer's, success and basically the company's success because we are relying on our customers being successful and happy to maintain our own life.
It's like a snowball effect that I believe will start with team happiness first, and that will transfer over to customer happiness and then company success.
Scott Tran:
You're also going to be attending the Oakland Summit. What are you looking forward to at the conference?
Andrea Silas:
I enjoy being around my peers, the people who speak the same language basically, who have the same struggles, the same wins, and the questions that need to be answered. Other types of leadership, other types of business thinking. It really is the community that I'm looking for.
Scott Tran:
Why is being there in person different than communicating online, even if it's the exact same people?
Andrea Silas:
Just sharing ideas that came out of a strange association. Like you start talking about something else and all of a sudden you pivot to another conversation. And that wouldn't happen online. Online here's my question, who can help me? And then it's about the people who happen to be online or happen to think that they have to do with something.
Being able to go over to someone or listen to another conversation that I might have missed because it's not something that I'm usually interested in and I'm finding something common with other people.
Let's be honest, we are human and we enjoy being in each other's stuff. I'm an introverted person, so I limit my human interactions, but I still need to be around other people and have those spontaneous conversations.
It really is about, being exposed to something that you might not seek out. But then you realize that I'm glad I didn't miss this.
Check out the video now featuring Andrea Silas our upcoming speaker for November's Leadership Summit