[SaaS & AI Series] Why AI Still Needs a Human Sherpa With Gil Makleff

Gil Makleff is the CEO and Co-founder of Sembly AI, a platform that helps teams turn meetings and conversations into actionable insights and work-ready outcomes. Under his leadership, Sembly AI has advanced tools for AI-assisted collaboration, workflow automation, and human-AI interaction. Before founding Sembly AI, Gil achieved two successful exits to Microsoft and Ernst & Young, where he later served as a Partner. With decades of experience building and scaling technology companies, he is known for championing human-centered AI that empowers people rather than replaces them.

Here’s a glimpse of what you’ll learn:

• [4:22] Gil Makleff shares the journey that led to founding Sembly AI
• [7:12] How Sembly AI is evolving beyond traditional meeting assistants
• [10:24] The “Sherpa” approach to keeping humans in control of AI
• [15:31] Sembly’s role across sales, delivery management, and meeting management
• [18:34] Rapid AI change and the future of agent-to-tool interaction
• [25:15] The emotional realities of selling a company and moving on
• [35:13] Leadership lessons for building empowered, high-performing teams

In this episode…

Technology is advancing rapidly, but as AI becomes more capable, how can businesses embrace automation without losing the human judgment that drives meaningful work? What does it take to keep people at the center of innovation?

In Gil Makleff’s view, successful AI is not about removing humans from work but helping them create better outcomes. He highlights the importance of designing AI systems that collaborate with humans, allowing professionals to maintain control while automating repetitive tasks. By transforming conversations into actionable business deliverables, AI can improve productivity without sacrificing critical thinking or decision-making. Gil also shares how his entrepreneurial journey shaped his approach to leadership, from navigating company acquisitions to building empowered, high-performing teams. He explains why values-driven decisions and a culture of trust are essential for long-term success in both business and technology.

In this episode of the Inspired Insider Podcast, host Dr. Jeremy Weisz sits down with Gil Makleff, CEO and Co-founder of Sembly AI, to discuss building human-centered AI that empowers professionals instead of replacing them. They explore the vision behind Sembly AI, the role of AI agents in transforming workplace productivity, and lessons from scaling and exiting technology companies. Gil also shares his leadership philosophy for building loyal, high-performing teams.

Resources mentioned in this episode:

• Gil Makleff on LinkedIn
• Sembly AI
• ChatGPT
• Notion
• Slack
• Microsoft
• Google
• Deel
• Stripe
• Ramp
• Microsoft 365
• Ernst and Young
• William Edwards Deming in the World of Six Sigma

Special mention(s):

• Dr. W. Edwards Deming

Related episodes:

• “[SaaS & AI Series] Smarter Way To Turn Meetings Into Outcomes With Artem Koren” on Inspired Insider Podcast
• “Automation Solutions with Wade Foster Founder of Zapier” on Inspired Insider Podcast
• “Pipedrive: Brain Surgery, Married, & Moved Company from Estonia to U.S. All at Once – with Urmas Purde [Inspiration]” on Inspired Insider Podcast
• “[SaaS Series] Revolutionizing Online Forms and Productivity With Aytekin Tank of Jotform” on Inspired Insider Podcast

Quotable moments:

• “Humans should stay in the driver’s seat, not become passengers to automation.”
• “Money isn’t the only driver behind those decisions. I’m happy with the choice I made.”
• “Empower people to use their strengths, and they’ll contribute more than you expect.”
• “I’ve never seen technology change this fast. Every month brings something new.”
• “We should work with AI, not be replaced by it.”

Action steps:

1. Keep humans at the center of AI workflows: Use AI to enhance decision-making and productivity while ensuring people guide, review, and shape the outcomes.
2. Empower teams by leveraging individual strengths: Create an environment where people feel valued and trusted to contribute, leading to stronger collaboration and higher performance.
3. Adopt new technologies to improve productivity: Regularly evaluate and implement AI and automation tools that streamline workflows and eliminate repetitive tasks.
4. Develop an AI-first learning mindset: Stay curious about emerging technologies and embrace new ways of working to maximize AI’s value.
5. Make values-driven leadership decisions: Prioritize long-term impact, team well-being, and integrity over short-term financial gains when navigating difficult business decisions.

Episode Transcript

Intro 00:15

You are listening to Inspired Insider with your host, Dr. Jeremy Weisz.

Dr. Jeremy Weisz 00:22

Dr. Jeremy Weisz here. I am the Founder of InspiredInsider.com, where I talk with inspirational entrepreneurs and leaders. Today is no different. I have Gil Makleff, one of the co-founders of Sembly AI. Actually, we had Gil, as you know, your co-founder, on the podcast. It’s a popular episode. People love it. And this may be the first time I’ve had two separate co-founders for one company. Gil has an amazing background and track record in entrepreneurship and business, so I’m looking forward to hearing your insights and thoughts. Gil.

Before I do, I always like to point out other episodes of the podcast people should check out, since this is part of the top SaaS and AI Series. Obviously, Artem’s was a good one. We’ll be biased on that. I had one of the co-founders of Zapier on. That was also an interesting episode, one of the co-founders of Pipedrive on. At the time, I think there were 10,000 customers. Now I think they’re over 100,000, which is interesting. The founder of Jotform talked about growing to 25 million users on a freemium model and how he thought about the freemium model when it came to a kind of product-led marketing situation, and so many, many more on InspiredInsider.com.

This episode is brought to you by Rise25. At Rise25, we help businesses connect to their dream relationships. We do that by one. We’re an easy button for a company to launch and run a podcast after doing it for 17 years. Plus, it’s one of the best things I do because I love giving and giving first, and I’m able to profile the people in companies I admire and share what they’re working on, to is gifting. So we will be done for you. Solution for companies’ gifting. They could send their clients, partners, or prospects, even staff, from a cultural perspective, and they just send us a list. We do everything else with addresses, and it’s not like, I guess, tchotchkes. Gil. I like getting food in the mail. So it’s all food, right?

Some people recommend not sending food because they eat it and it disappears. But we send something every 3 to 4 months for like four years. So it’s like a constant reminder, like, hey, I’m here. I appreciate you. Enjoy. So check out more Rise25.com or email us at [email protected].

I’m super, you know, excited to introduce Gil. He had before Sembly AI. He had two previous exits from Microsoft. He’s a former EY partner. And he really believes in keeping humans in the driver’s seat of a genetic technology. And as I mentioned, the Co-founder of Sembly AI, they turn conversations into client-ready deliverables. And they’re really a genetic solution for professional services, high-impact teams who don’t want to waste time and get stuff done quicker. Right. So, Gil, thanks for joining me. You’re going to add something to that.

Gil Makleff: 03:22

Thank you. I’m excited to be here, especially as your first double co-founder.

Dr. Jeremy Weisz: 03:28

Exactly. So, because I’ve had people together but not separate before.

Gil Makleff: 03:33

Yeah.

Dr. Jeremy Weisz: 03:34

There are different journeys here with you and Adam as well.

Gil Makleff: 03:36

That’s right, that’s right. We have different journeys, although we have known each other for so long. But and we, we support each other. But yeah, I’m really lucky to have a partner like him.

Dr. Jeremy Weisz: 03:46

How did you meet Artem?

Gil Makleff: 03:49

Artem. I met him at UMT, my previous company, and we clicked immediately, and it was just, you know, we got to know each other over the years and worked together. And yeah, he ran major, major transformations. And we interacted together on these, on these big projects. So yeah.

Dr. Jeremy Weisz: 04:13

How did it work at one point? Did you go to him? Did he go to you and say, we should start something?

Gil Makleff: 04:22

I think it was a phone call. I was a partner at EY, and I think. I might have reached out to him. As you know, we were very close. So I was at his NYU. We’re both Columbia and NYU grads. So I went to the NYU graduation, and we got close. So it was a phone call, probably that we said, you know what he was traveling at the time. I think he might have been even in Japan or something. And I said, you know, all right, let’s do something. And we got together in a whiteboard room and started. I still have the picture of that whiteboard with about a thousand different names, ideas, and things like that.

Dr. Jeremy Weisz: 05:07

So I’m always curious about that, Gil. Because, like you could have done, like you said, a thousand different things. What else was on that board that I don’t know if you remember, because it’s been a while, but what was on that board that you’re like, this is a great idea, and you’re almost going to pull the trigger, but it didn’t end up happening because you went in a different direction.

Gil Makleff: 05:28

No, it was more of a conversation between Artem and me about our work in the past and how we could make it better. And a fundamental part of our work was meetings, and we all had been in meetings where people come in, and they leave, and nothing happens in between. And there’s no track record, there’s no follow-up. And, you know, basically nothing happens, especially in large corporations where people know how to do that really well. So yeah, so we, we kind of honed in on this area and it took us a, you.

Dr. Jeremy Weisz: 06:05

Knew it was going to be something around meetings. Like those ideas were around.

Gil Makleff: 06:08

It was around it, you know, it was around the initial part. It took us a long time to leave this meeting. We’re away from the meeting content concept today. We moved. And, you know, very soon, meetings are going to be a feature of our product. Not really necessary if you don’t want them to. Good. Listen, we have an amazing platform where you get great information, which allows you to be more contextually accurate to deliver results. What are the results there? As you said very eloquently at the beginning, deliverables, we built a deliverable for you, as you would build it, and we allowed you to go over that deliverable. And I think we should talk about that idea because I think that, yeah, anyhow, I don’t.

Dr. Jeremy Weisz: 06:58

What do you mean by getting? Tell me, talk about getting away from meetings and what that means. I’m going to pull up your website so people can see, because, like Sembly.ai, you can check everything out. If you look at the product when you say, we’re getting away from meetings, tell me about that.

Gil Makleff: 07:12

It’s not getting away. It’s it’s meetings have become more and more commoditized. Being an agent in a meeting. There are thousands of them. There are so many. So we, we like to think that we’ve always been at the forefront of innovation. And we came out with new features and capabilities, always ahead of the pack. And we’re consistently doing that now. We’re building capabilities that are coming out at the end of the summer, which are again, groundbreaking and powerful for working teams in professional services companies. So yeah, so that’s, that’s what we’re doing. We’re enabling capabilities, but we’re thinking about this in a way that we think again is a little different than the marketplace, because a genetic technology allows you to automate everything.

There is a huge degree of disintermediation in the marketplace, which means things are done differently. Users are addressing simple processes. They have always done things in a different way. And this will happen more and more and more. And, you know, we spoke a little bit about this before, but the market is looking very, very different in terms of a simple example, which is how people get help online. For example, you used to have to go to a vendor, any vendor name that you can think of, and get the support team of that vendor, call, and hope you get a human who knows what they’re talking about to help you solve a significant problem. Later, you could get the support from an automated app and and get the support. Today, we went to another phase where you’re using the LLMs of your choice, taking a picture.

Because of the multimodal approach that we have today, we can just take a picture, and the LLMs sees exactly what you’re seeing and helps you navigate those issues. So simple that it calms you down. It makes it easy. Everything is easy. So I think we’re living in a new world where. Humans are learning to do things differently. What’s with the danger of that is that we can get to a point where, slowly, the human doesn’t have to do anything, and it’s possible to do it today, where there’s very little to do for humans. And I think humans need to defend against this process. We need to enable human participation in workflows in corporate environments. We have to fight to enable this. And at Sembly AI, we’re really aware of this, and we’re building capabilities to address this issue. And I could talk about that.

Dr. Jeremy Weisz: 10:20

Yeah, talk about that. I don’t know if you’re referring to the Sherpa interface.

Gil Makleff: 10:24

Yeah, yeah. We’re building a Sherpa. I’m calling it a Sherpa interface. It could get other names because.

Dr. Jeremy Weisz: 10:30

So like, let me ask you a question real quick. Sherpa interface piece. So eventually, like, will people see that show up in like under product, or where will they be able to kind of see that when it comes out? Because I see features and there’s automations, right?

Gil Makleff: 10:44

It’ll be front a frontal in in the August time frame. You’ll see it, you’ll see it at the time. It’ll be there. I call it Sherpa because a Sherpa has human experience, has knowledge, has human expertise that we’re in danger of losing if we automate everything and rely totally on LLMs. And that’s kind of a fear that we need to address in a way that makes collaboration possible.

And you can tell me, oh, but you know, at the end of the day, you know, LLMs agents are going to do everything. And that’s a correct argument in a sense. Logically, it’s correct. In the future, agents will probably do most things.

But I think if we build technologies that are like the Sherpa interface, where the human is the Sherpa Guiding. Think of holding a steering wheel in a car. That’s kind of the best analogy I can think of. The human helps determine whether you are making a right turn or a left turn. And getting to the right destination is also partially a human success. So the human not only provides input, but determines the result, which is really important because we need to feel empowered. We don’t want an agent that’s so good that does everything on their own. We want to be a part of that process. And how do you do that? You know, that’s the question.

Dr. Jeremy Weisz: 12:24

And you tell me, I don’t know.

Gil Makleff: 12:25

You do that by enabling a Sherpa interface, which allows for exhaustive questions and answers. And how would you modify what I have already? Well, let’s take this right turn and implement something here, or let’s do something. So enabling the dialogue in a way that’s thoughtful, in an environment that feels like you’re impacting the result in a meaningful way, is going to create end results that are not only correct and accurate, but are exactly what the human intended to deliver. And that’s something that we think is going to be a path that users are going to want in order to achieve success generally, and in the future, in the foreseeable future.

Dr. Jeremy Weisz: 13:17

Walk me through that for a second, like maybe a use case that you’re thinking of. And then how would it work?

Gil Makleff: 13:24

Sure. So let’s think of. We have quite a few use cases we’re working on, but let’s say a recruiting Workflow. So what we’re we’re enabling in the technology and it exists today partially already is a framing of the type of questions you would like to ask in the detail that you would like to ask, and a result of a review, let’s say, of multiple interviews of a specific resource, but allowing a back and forth movement between those two, the result that you get and enabling more query and more query, and then making decisions based on results that are more meaningful to you in that way.

Dr. Jeremy Weisz: 14:23

It’s interesting to look at this, Gil, because I like seeing one. And if you’re listening to the audio, we’re at the Sembly AI and the use cases, because like, how should I be using things in general and for what? Even seeing with the product because it’s a mindset shift, I think, for someone like you, for anyone to take a picture and then push it through the LLMs, I mean, that’s just not my first instinct to do, right? I’m kind of used to behaving like go, go to Google. Oh, wait, I should be going, just talking to ChatGPT. I don’t even think of uploading a picture.

I’m thinking more like an even speech-to-text type of thing. And yours is way easier, actually, as you say that, but I like to look at your product page and think, okay, what are automations that I am not automating right now that I should be right now? Because you have analytics, collaboration, CRM, custom integrations, and all these things. Same thing with deliverables and with use cases. I mean, all this is, I think, instructive for any business, in my opinion, to look and see how you could be using automations, AI, and everything like that, you know.

Gil Makleff: 15:31

So what we’re building is across the board for professional services. It starts in the sales process, then moves to delivery management, including meeting management. Think of all those areas. So imagine meeting a customer in the afternoon and describing an offering that you want to sell them. Think about that. And using the LM or using Sembly AI or Sembly 3.0 to produce a document that is branded with the customer environment. Touch and feel, logos, colors, fonts, etc., and the context of the information that you know, Sembly AI not only has information about what was discussed in meetings, etc., but also is like a fabric in the environment.

It’s a platform that connects to many, many applications that fabric is becoming a two-way. In the past, it used to be a one-way direction into applications. Today, it’s being built into a two-way environment. So think of Sembly AI as getting information from systems of record and connecting into live meeting environments, and producing end results. If you point it to the website of the potential customer, Sembly AI knows how to do all that natively. So it’ll take the information it’ll take, it’ll marry it with live information from meetings, and it’ll create a deliverable that looks and feels very authentic and powerful. And it can do that, obviously, in minutes.

Dr. Jeremy Weisz: 17:24

I’m curious what tech and apps you are using daily? And it could be personal or business.

Gil Makleff: 17:35

We everything. I mean. Notion. Slack. Microsoft. Tools. Google. Tools. We obviously use many, many tools.

Dr. Jeremy Weisz: 17:47

Yeah.

[Continue to Page 2]