This website requires JavaScript.

Talking efficiency and automation as a customer service agent, with Hosam Hassan

Customer service27 MIN READSep 27, 2023

Last episode, we defined AI in human terms. Now, let’s talk about the easiest ways to start using these technologies.

Our guest Hosam Hassan gives excellent advice to professionals and founders far & wide. His success began with discovering ways to streamline workflows through automation, and he’s eager to impart this knowledge to others. 

Listen in to discover:

  • Why agents should be excited about AI,
  • How attitudes towards automation are changing,
  • Practical strategies for AI implementation in customer support teams.

Grace Cartwright is your host – she tackles thought leadership content at Klaus and loves talking to anyone making waves in the CX space. 

Hosam Hassan is your guest – a customer service agent turned entrepreneur, he now helps modern businesses with tech-solutions as Director of CXT Strategy at PartnerHero.

You can also read the podcast transcript in full below, and we also recommend:

  1. Following Hosam on LinkedIn for regular, applicable advice.
  2. Listening to episode 1 (if you haven’t already!) to understand what some of these terms mean. 

Transcript: ​​S03E02 

S03E02 Quality Conversations with Hosam Hassan

Grace: Welcome to the third series of quality conversations with Klaus. I’m Grace and in this series we’ll be tackling a matter on everyone’s minds: AI. but more specifically how it relates to customer service careers. 

In the last episode we turned AI inside out and upside down to figure out what these new technologies actually mean. This episode, we’re looking at the easy ways it can help you. 

Hossam Hassan is our guest. He is currently Director of Customer Experience Transformation Strategy for PartnerHero. Hossam himself started out as a customer service agent and he knows firsthand how using tools wisely can give you a huge career boost – having gone from agent, to entrepreneur, to consultant.

You’ll learn the best approach to adding new tools, as well as his predictions on emerging trends. Let’s get into it.

Grace: Hello, and we are onto podcast number two in our series about customer service careers and AI. I have with me today Hosam, who is a customer support agent turned entrepreneur. 

And he recently shared with me that he is very optimistic about the role that AI and automation will play in the future of cx. And I’m really excited to hear why exactly he is so optimistic.

Hosam: Yeah, Grace, thanks for having me. Always a pleasure to talk to someone at Klaus, all great people, and a great product as well. So, it’s my pleasure to be here. 

But, now I’m the Director of Strategy over at PartnerHero, and one of my main goals is to figure out how we could take a traditional outsourcing model and implement AI with it. So, because I have a little bit of experience working with AI when I was over at Ultimate, I know the power it could have over automating tasks for agents, taking the workload off of them, and actually helping them become more efficient, and more consistent with the replies they sent.

Grace: Absolutely. And do you think that’s something the agents themselves are very keen then to take on board?

Hosam: Yeah, actually when I was an agent myself a few years back, I was always looking for ways to automate my own workflow. Sometimes you get stuck in this loop where you’re doing the same type of ticket over and over and it’s really draining, right? 

And even the problems that Klaus tackles, like QA, imagine a QA Analyst just consistently reading tickets and looking for the same things. The job could get very boring. So if you have the tools to help you automate those manual tasks, then the tasks that you are actually working on are unique and allow you to use different parts of your skillset. So I definitely think it’s an advantage for the agents and they should be very excited about it.

Grace: And in that respect, have you seen attitudes evolve? Because there’s, of course, conflicting opinions at the moment about where AI is right now, where it’s going and what people can expect. And there is a certain amount of trepidation, I think, with people wondering how their jobs are going to change.

Do you think you’ve seen attitudes change quite a lot then in your career?

Hosam: Yeah, I mean it’s definitely still a sensitive subject. 

It’s kind of when Lyft and Uber came out and they said, Hey, we’re gonna take over all the cab driver jobs. And they did take some. But what ended up happening was the cab companies evolved. They either adopted Uber or Lyft, and brought in that new technology.

Here in New York, the cab drivers have their own app as well called Curb, and they’ve adopted that new way of hailing a ride through your phone, as opposed to, you know, putting your hand out and possibly risking an injury or something.

But we’re going through the same thing with AI, and people are worried about their jobs. What I will say is there’s gonna be a transition period where some people are gonna be resistant. But then when you realize that the technology’s there to help you as opposed to replace you, that’s when you could really take advantage of it.

All of the tools that I’ve seen today could only do so much for the workflow. There still needs to be a human in the loop, right? You can’t remove the human at this point. Maybe someday, five, ten years from now, processes will be so automated that humans will get to work on other things.

But for now, the human is still required to close the loop on most tasks. 

Grace: Absolutely. That’s a hundred percent what we espouse at Klaus as well. That, you know, our latest AI feature, AutoQA, is used alongside manual reviews. It’s not going to completely replace the analytical skills and the soft skills that an actual person, who knows the business, who knows the customers better, is going to be able to apply. 

I love your analogy actually about Uber. I’m based in Prague and that’s exactly what’s happened here.There’s Liftago, which is the local app, and it just works great for everyone. Yet five years ago, I think there was the same fear of, you know, these taxis thinking that this is our jobs being taken. 

Whereas I think there needs to be a certain amount of reasoning that needs to come instead of this fear mongering that often the media seems to purport about how AI is going to revolutionize everything, including humanity as it stands. It is fear mongering, I feel.

Hosam: Right. Yeah. We can’t forget that AI is built on historical data sets. It’s based off the past and it really can’t do any predicting, other than what has already happened. So we still need to generate that data and plug it into the AI for the AI to be effective. 

We’re not at the point where it’s self-learning and creating these future predictions based on data that hasn’t been inputted into it. 

Grace: Exactly. It’s based on us, essentially. It’s much faster, absolutely – and in some circumstances, not necessarily smarter, but it can reach a more direct conclusion. But it doesn’t necessarily have something that humans cannot possess.

Hosam: Exactly. 

Grace: With that said, you’ve worked for many companies that aim to help customer service teams get better and get more efficient. What would you say are the best practices that you can offer and that will be still relevant going forward?

Hosam: Like I mentioned, I’ve been a support agent myself. I’ve also been a support manager. Then eventually went into support operations where I was looking for these efficiencies and these automation potentials. And I always start with – it might sound lazy – but always start with the easy work.

Find the use case that is gonna be the easiest to automate and start working on that one. Maybe it’s your highest volume one, maybe it’s not. But if you start with the small things or the easier tasks and automate those, you’ll get into a rhythm and eventually get to the complex things.

But if you start with the big, hairy, super complex situations, you might get discouraged from automation and you might actually mess up a few things before you nail it. So I would go after the really easy things like auto routing your tickets, for example, auto tagging them, auto triaging, auto routing, sending auto responses to very basic questions like, Hey, what is your return policy, for example. That doesn’t change, that’s gonna be the same. As opposed to, I’d like to make a return, I want to change it to this size, and so on and so forth. 

And then the second tip is, like anything else, document your work so that way you have a record of what you’ve changed, in case you need to revert an automation and go back to the manual process. So you could start with a small segment to see how it does, and then, if it’s successful, continue and document along the way.

Grace: I’m curious, have there been many circumstances that you see that people use this on a personal level rather than just a team level? Because there’s many, many, many, teams that are, of course, starting to adopt these tools. But there are some managers out there, or team leads, or agents even who are listening to this who don’t necessarily know that they have that backing on a more departmental basis. Whereas they want to do things that would affect their own work personally.

Hosam: Yeah, so we have that happening on my team right now. My partner in crime, his name is Craig – I’m Director of Strategy, he’s director of delivery. So he makes sure the implementations go out super smoothly. He figured that it wasn’t the best use of his time to take notes during meetings.

He went out on his own, found a tool that takes notes, and now is sharing that with the rest of the team. So at the end of the call, we have this amazing script written by AI of exactly what happened on the call and instead of him asking our boss or asking finance, he just implemented the free version. And I think eventually bought it himself to prove the concept. And now that we’re all invested, we want to get that tool for the team. 

So I think as an individual contributor, as an agent, manager, or leader, if you find opportunities and you’re afraid of friction, try it out on your own. See how it works – if, obviously, if it passes your security things at work!

Go ahead and implement automation on a small scale, prove the concept and then try to sell it to the team. 

Grace: And no one can argue with evidence, right, if it’s something that’s gonna make you better at what you do? Team notes is one that I was talking about even yesterday, because it often gets to the end of a team call and then it’s like, wait, is someone writing this down? It’s so basic and yet so easy.

Hosam: Yeah. Or forgetting to record, right?

Grace: That gave me a sudden fear. But don’t worry, I clicked the record button before this!

And do you think there’s any danger? What people, especially in support, are really concerned about is that there will be this disconnection with the human touch with service and personalization. Do you find that there’s any issues with over-automating or relying on technology too heavily?

Hosam: Yeah, I wrote about this a few times on LinkedIn. 

There’s certain parts of your life or certain parts of your day where you want to interact with a human. Like when you’re talking to your parents, talking to your siblings, talking to your family, going to the hospital, there’s an emergency, right? That’s when you want to interact with a human. 

But if I’m ordering food, if I’m returning something, if I’m buying something, if it’s transactional. Personally, I don’t want to really talk to somebody and I think a lot of consumers are the same way. So make sure you implement the human touch when it’s necessary, but you could feel comfortable to remove it when it’s not right.

So things that are transactional, I would say those are lower on the empathy scale than dealing with an emergency situation, an incident. A high value product, if you’re selling products that are tens of thousands, hundreds of thousands of dollars, that’s when you’ll need to implement the human touch.

I think it’s case by case, but that’s my take on it. Save the human touch for when it’s really important, and use good judgment when it’s not.

Grace: That’s a good point. I think my query is maybe because it depends on the customer. That point is different for different people, right?

I completely agree that, in most circumstances, I don’t want to get to the point where I have to speak to someone. I want it to be fixed as quickly as possible and preferably in a way that I’m able to independently figure out – or with minimal effort from myself.

So we’re seeing a lot of these smart chatbots appear. Do you think that they are smart enough to know exactly how much, or maybe exactly when to offer over to a human? It’s called the escape hatch, right? When it’s that time is, before it’s too late, almost.

Hosam: I would say build as many workflows that are automated as possible. But always put that escape route right in front of the customer from the beginning, so they could escalate to the agent whenever they want. Don’t bury that. 

There should always be an escape because some customers, like myself in certain situations, I’ll just continue down the path with the bot and if I get it resolved, I get it resolved. If not, then I’m like, you know what? I feel like this is a special use case. Let me talk to an agent and I want to be able to just divert immediately. So, I would say put it front and center right from the beginning. No need to hide it. 

And I do want to add one more note. If we flip this and think about the agent, the agent is talking to people all day. Maybe eight hours a day, ten hours a day, they could be doing phone support, chat support, email support. Think of how exhausting that is for the agent. By implementing AI, you’re allowing an agent to save some of their empathy for when they go home and they want to interact with people. 

I know when I was an agent and I spent eight to ten hours a day talking on the phone, when I left, I didn’t want to even have those social interactions cause I was just depleted. 

These tools are not only helping the customer, but they’re helping the agent do more focus work. They could be writing help desk articles. They could be building more automations. They could become conversation designers and build these workflows. They could do QA. So they could be doing a lot more proactive things than just talking to people all day. And it doesn’t allow them to completely be depleted by the time they go home and want to have these social interactions. 

Grace: I think that obviously is beneficial on an individual basis, but it’s also beneficial on a company basis. Turnover in the customer service industry is high, especially with agents who are talking to frustrated customers day in, day out. There’s only so long that you have the energy to sustain that, and to sustain your own soft skills or your own reserves for empathy, I think. 

So on a long-term basis for the company, that’s only good because it means your employee retention is higher. And you have a team of experts rather than a team of newly onboarded agents.

Hosam: Exactly.

Grace: Are there any companies that you see really excelling in this, that you are inspired by or that you are proud to have worked with?

Hosam: I think e-commerce in general is accelerating the quickest. They’re always ahead of the curve, especially D2C brands, because a lot of their requests are repetitive and transactional. 

I think it was On Running – I bought shoes from them and wanted to return it, And I didn’t really have to talk with anybody. I just went through their automated workflow and got the return, got a return label and didn’t really have to do anything other than ship it out. So I think e-commerce in general is winning at this.

But B2B will likely catch up in the next few years.

Grace: Is there a reason that B2B has been more behind?

Is it just simply due to more complex queries?

Hosam: Yeah, because when you think about it, B2B is gonna be SaaS products and technology. They’re so unique and they’re so new that it’s not like just connecting to a Shopify store or your e-commerce platform. Every company has their own proprietary software, their own proprietary backend, their own APIs.

So you can’t say yeah, let me just easily plug into this tool and automate a bunch of tasks. There are companies that have a very complicated password reset process, right? Because of their technical debt, how they built their infrastructure 10 years ago is not allowing them to automate that. So with B2B and software, it’s gonna be a while before they have that same accessibility as an e-commerce company.

Grace: Yes. I’m almost curious if in the near future we’re gonna see a lot of frustration with people working in bigger companies because of that very reason. Because of the security and the compliance issues with introducing new emerging technologies and tools – which maybe the smaller, newer companies are able to snap up and adapt to a bit easier. It’ll be interesting to see how everything plays out. 

What do you think are emerging customer service trends that you see, that maybe you wouldn’t have predicted in the past?

Hosam: Obviously what’s been going on with ChatGPT is incredible. We’re seeing the development of tools accelerating at a crazy rate. 

Craig and I are working on a project where we’re doing discovery on a bunch of different tools so we could understand the AI landscape. And once we feel like we have a complete list, we’ll go away for the weekend, we’ll come back and there’s like 10 new tools on the market!

So I think the trend is that a lot of people are gonna build a lot of single feature tools. Tools that just do one thing, but then there’s gonna be a point of consolidation where they’re either gonna get acquired, or get merged, or they’re gonna shut down.

Mike, CEO of Ada, dropped this in LinkedIn the other day. He said that most tools are gonna be AI native at some point, where AI is just part of your technology. It’s not the special thing that you gotta upcharge for – it’s just like embedded. 

And I think Klaus is doing exactly that, right? It’s become native to your product, it’s not a strange outlying feature. 

So I think that’s the prediction. It’s gonna be consolidation and, and the, the technology becoming embedded instead of just this like standalone feature.

Grace: I completely agree. I think it’s difficult though. 

What our challenge is at the moment, at Klaus, is keeping all of our customers and our prospective customers up to date with the changes that we’re making. Because if you have a strong data team, that’s fantastic. They can now move far quicker than they could even twelve months ago. But educating people on how best to use that, and any best practices? It takes time. It takes time for people to get used to change, also, because we are creatures of habit. I think that’s gonna be one of the big challenges going forward. 

You said that you and your partner made this big list – that is an ambitious endeavor. You think that more and more companies will consolidate. 

Do you have any tips for making sure that you stay up to date with what’s going on in AI? How to upskill? And how to basically stay on track with this pace of change?

Hosam: Yeah, we’re doing it because we’re in the space and we wanna make sure we deliver the best solutions for the problems our customers bring to us. But if you’re a support leader, or anybody at a company, and your role is to focus on one department: focus on the problems that you need to solve. Don’t get overwhelmed by the tools.

If I’m doing QA, I know my job is to solve QA-related problems, I’m not gonna go look at tools that are unrelated to QA. It doesn’t make sense. So I’m not gonna go look at a sales AI tool and be like wait, we should implement this at our company. 

Focus on what problems you’re solving today, and things that you’re doing manually today, and see if there’s a tool that will assist with that problem. If there isn’t, don’t get overwhelmed by the tools because they can get extremely overwhelming if you just look at everything that’s out there today.

Grace: That’s very, very good advice. Sometimes with all of the noise that’s going on, it can be tough to channel what’s actually gonna be good for you. So starting out with the problem is a perfect approach, I think. 

And this is a really tough question probably, but what do you predict for customer service for the future? 

We talked about, maybe in a decade then we’ll see a much more seismic shift with how roles are played out and how people work with technology. But what do you maybe see in the nearer future for people who are job hunting?

Hosam: Again, I’m gonna quote Mike, CEO of Ada, cuz I really like this quote. He said that AI is gonna be native. And the second piece to it is that instead of AI augmenting the agent, the agent will augment AI. 

So AI will do a bulk of the work and then the agent will come in and wrap things up. As opposed to how we’re doing today – the agent does most of the work, and then AI just does a piece of it. So in the future, we’re gonna just get used to AI dealing with most of the complex backend things that we’re doing today. Things like jumping around from tool to tool, processing a return here, and then jumping into here to document it, and then jumping into another tool to let the customer know … AI will handle all of that.

We might just be sitting in front of a computer and saying okay, everything looks good, send this out. We’re gonna be able to put the finishing touches on it from that perspective. And then we’re also gonna become the builders of those workflows. So we’re gonna be thinking about how we automate from day one instead of, let’s automate a manual process that we’ve had for years. We would be starting from scratch and building with automation and mind again, making sure that AI is native to that workflow. And then when we can’t automate, that’s when we’ll get the human in the loop. 

He put it that way. I agree with that sentiment, I believe that’s gonna be the future.

I think for those people that are worried about losing their jobs and so on, although there will be a small wave of that, I think getting ahead of that curve, you could get into conversation design. That’s where you’re the one building and designing bots. Or getting into becoming an AI trainer where you’re training the AI on data, helping it not get confused between different inquiries. A chatbot builder is very close to conversation design. 

Again, QA is still gonna be something that humans need to be involved in. Even if you’re running QA through a bot or through AI, a human will need to QA to make sure that it’s working properly.

So there are many emerging roles that customer support agents could start looking at today. If they want to upskill, maybe change their career or just prepare for the next five to ten years.

Grace: You’ve already answered my next question, which is what are the practical tips? But that’s perfect really, to more look at how you do things in a strategic sense, rather than breaking everything down into these tiny things.

What we’ll eventually be able to do is be more architects of how conversations go and how to optimize them. Hopefully, I mean this is our hope with Klaus, is that we just understand our customers better. Because we have the time to do that, and we do have the analytical skill to do that, because AI will hopefully be taking over and automation will take over the more mundane stuff.

Hosam: I mean, as long as you have customers in this world, you’ll need customer support people. In one way or another, you can’t completely hand that off. So I’m optimistic about the future.

Grace: You’ve now answered my last question: is AI coming for our jobs?

Hosam: Like we talked about it a little bit before, we used Uber and Lyft as an example. I’ll just throw another one out there that I just thought of…

Back in the day, we had mathematicians doing pen and paper. They moved on to calculators, now computers, yet we still need mathematicians. There’s no point where we said we no longer need mathematicians because we have calculators and computers that could do all of this. 

So, AI isn’t necessarily coming to take your job, it’s coming to take care of the manual tasks. So your boring work is almost gone. 

Well, let me not say all your boring work, but the repetitive work is gone. You may have some boring, unique tasks to take on, but they’re still unique and you won’t just be a machine clicking through, cuz that’s what customer support sometimes feels like. I know firsthand. Sometimes you just feel like you’re just going through a process, filling a seat, and not allowed to be strategic. But as the shift comes, I think a lot more people are gonna get into the proactive side of customer support versus the reactive side. 

I think there’ll be a small wave of jobs being lost, like any shift in technology. But not to the point of what everybody’s thinking, of this mass massive like layoff coming up.

Grace: I agree. Also, as you’ve discussed, there will be other – not necessarily new jobs – but roles will evolve to adapt to how the technologies will evolve well. 

Much like customer service used to be sitting and picking up a phone. Now it looks like a completely different industry nowadays. Just like AI will probably instigate another way for us to evolve in how we communicate, I think.

Hosam: Yeah, and if the job itself, let’s say the customer support role, is no longer needed in 10 to 20 years, the people working in customer support are so well equipped to work in any other department at the company. Like, someone from customer support could easily transition into sales, easily transition into product, easily transition into marketing. 

And I say that because they’re the closest to the product. They know the ins-and-outs of the product, and they could say, okay, I know the product so well. I know the flaws of this product. Let me go join the product team as a product manager, and help them discover these issues and fix those problems. 

So I’m not saying everybody in support will change their roles, but there are opportunities to use those skill sets in different departments and different roles, outside of the emerging roles that we already talked about.

Grace: Yes. In our last podcast series, we actually spoke to Inna Grynova, who is Head of Support at She spoke about how she’s ex-Google, and they talked about how hiring overqualified people was always their strategy. Because you are essentially starting someone off in your company who is going to learn the most, because they learn face-to-face with customers. They are handling the product day in, day out. So they are the perfect people to be able to progress and and feed that information into other departments, so that they can become more customer centric. 

That’s hopefully a positive thing for customer service going forward too, you know, for many companies anyway – the better ones – the shift from customer service is this cost center to a profit center. Hopefully if AI and automation helps these departments become even better at what they do, then they will become even more useful as a voice for the company and as an ear for the company.

Hosam: Exactly. When you’re in support, it’s almost as if you’re doing customer research. You’re getting customer inquiries all day. You’re understanding the customer, their problems. It’s very much in line with what marketing is looking for, they want all that information. And if those two departments align, or if a customer support agent moves into marketing, they already have access to that information, which is great for the entire company.

Grace: It’s actually something that we do at Klaus – and what we encourage other companies to do – is for everyone from every department to read through support interactions so that they understand what customers are saying. The pain points and also the highlights. Cuz that’s the nice part of it too.

Is there anything that we didn’t talk about that you’d like to mention or discuss? 

Hosam: I think we covered it all. I’m just looking forward to what’s gonna happen in the next year or two. I think we’re the same energy that was felt when the internet came out, or when the internet became popular. I feel like we’re going through that as well. And some people maybe younger than myself – my sister’s age – didn’t get to experience that. And now they’re seeing that firsthand – a big shift in the world, and glad to be part of it. 

Grace: I completely agree. But I think the fear is not unfounded. On the other hand, I think hopefully people will find more and more reason to feel positive because there’s so much good that we can do. And not just customer service, but on a much bigger sphere, hopefully. 

I always just say I can’t wait for AI to do my laundry, and then I’ll be really happy 

Hosam: I mean, there’s a couple of people that I’ve heard say oh my God, AI’s gonna come replace all our jobs and we’ll all be out of jobs

Well, if we just wait for AI to do everything, we’ll all just end up retiring and hanging out at the beach and letting AI handle all of the work for us. You gotta think of what the positive outcome could be as well. 

Grace: I won’t bet on us all just being on a beach anytime completely soon. But we can dream right? By the time I wanna retire, maybe I can dream for that.

Hosam: Exactly.

Grace: Well, thank you so, so much for chatting. This has been great.

Never miss an update

By subscribing you agree to Klaus' Privacy Policy and would like to get educational content to your email.