Dr. Jeremy Weisz: 15:21
Yeah. There’s a lot this can do. And what’s interesting, it sounds like people I don’t know if people are using it from like an account management perspective, like I know, like we use ClickUp as our project management. So I imagine you could have a call with the client and then automatically pipe it in somewhere into notes so that it doesn’t have to be like recorded inside of ClickUp or highlighted or something like that.
Artem Koren: 15:50
So you have it open right now. If you kind of mouse over where you just were. And on the right side, you have automations and you can actually click into project management. I think you’ll, you should. The last entry there. Yeah. So yeah, exactly. Like you said. So Sembly automatically detects items like tasks and, and there’s ClickUp and it’s, it’s a 1 to 3 process to set up that connection. So you, you just connect from as into ClickUp will ask you to authenticate into ClickUp. You’ll say yes. And then it’s a one, two, three setup. And the one, two, three is you tell it first what items you want to send. Is it tasks? Is it notes? Is it transcript?
The second thing you’ll tell it is for which meetings do you want to send it? Do you want to send it for every meeting that comes through for you on Sembly? Or do you want to set it to like some subset of meetings or maybe like, you know, there’s many options, right? So you can filter down like only specific meetings should be routed to the tasks from only specific meetings should be routed to ClickUp.
And then the third piece is in some cases. So like ClickUp, I think is one of those cases. Yeah. There’s like a project destination or a board, a target board that you need to put it on. So that would be the third step. The third step may or may not be relevant for this integration you’re setting up. That’s it. It’s that easy. It’s 123 done. And once that happens, every meeting that flows through Sembly will automatically update ClickUp for you and you’re set.
Dr. Jeremy Weisz: 17:27
What I like about looking through this Artem is like, people can just get an idea of what they can actually automate that they probably aren’t. Right. And so if you look at this, I encourage people to check out Sembly.AI and you can go to if you do the use cases and then look at the automations and look at some of the stuff, you can actually automate that. Maybe it’s manual right now, right? And so we have CRM, we have collaboration analytics, custom integration.
I mean, I want to go through every single one of these. We don’t have time, but I’m going to have to do it on my own time and look through these. And what, what should we be automating and how can Sembly help with that? Right. And so document management, personal productivity, project management, all this stuff could save I don’t know, I don’t know what’s been reported, Artem. If someone can save, you know, across our team per person, maybe they save one, two, three, four, five hours if they have 20 people, that could be who knows, 100 hours of savings.
What have you heard from some of the clients? Maybe from the project management standpoint, you have a company in France. Talk about how they’re using it, how are they saving time?
Artem Koren: 18:37
Yeah. So well, first of all, we did a study across our major customers to calculate the ROI of Sembly. And what we found is on a 100 user basis. So like every 100 seats on Sembly that your organization uses, and then you can just kind of scale that up or down for your team. For every 100 seats, the ROI is north of 2.5 million a year. Now, Sembly costs a tiny, tiny fraction of that tiny fraction of that. You know, for like our most popular, the pro plan, it’s basically $20 per user per month. If you have a large organization like, you know, starting to get towards the enterprise side, you know, it goes up to 30 bucks a user. So figure at 20 bucks a user a month per 100 users, you know, that’s 20 times 20K times 12. Sorry, 10. Sorry it’s 2K times. So it’s like 20 4k per year. Your return is over 2.5 million.
Now you’re going to say that’s a crazy number. Like there’s no way it’s possible. But when you think about it’s not just the savings of time, like, oh, we saved you like hours writing meeting notes or like we saved you hours entering tasks. It’s also the time of like that you spend generating project plans, generating task reports. Semblian knows about all your tasks. You can ask Semblian, like for this project, what are the tasks that are due for me? It will tell you it’s always up to date. It’s always up to date in the latest meetings.
So there’s kind of an aura of value that it delivers. So the ROI is just phenomenal as far as the large customer. Yes. So one of the very large kind of project and PMO consultancies that works with the, you know, the largest companies like the largest airplane makers in the world. The largest car makers in the world. They provide a layer of project controls for those organizations. So for them, project management is their lifeblood. They love Sembly because it’s built with PM in mind. It’s built with professional services in mind.
So we understand the project construct. We understand the task contract. I’ll give you. It’s like it’s a little bit of a techie example, but I think it’s important. Like in many other products, tasks are like a text block, right? So there’s like a meeting notes block, that’s some text. And then there’s a tasks block. That’s some text. For Sembly a task as an object. And it has a task name, a description, it has a kind of a task. So if it’s like a, you know, kind of a small, medium, large. It has an assigner, assignee, a due date. So it’s really like PM-oriented, like it’s got all those elements to it.
And that’s why we can manage those objects individually. Like for example, sending them to ClickUp or sending them to Asana or sending them to Monday, wherever you may work. So they love that piece of it. They love that connectivity. And they also love the fact that they can collate information from a very wide range of teams very easily. I’ll give you a fantastic example. I actually really love this example from this particular customer out of France, which is because they service very large customers. Their project renewals start about two years before the renewal is due. You know, this is, you know, we’re getting big enterprise world. This is how things rock.
So two years ahead of the renewal, they start putting together the renewal paperwork. What this means is there’s teams in different countries that are responsible for dozens and dozens of work streams. Hundreds and thousands of work streams collectively have to put together a cohesive renewal proposal over some period of time. To manage something like this is incredibly complex, but to be able to status that up to the executive team and you can imagine, you know, you know, you have a client. Like, you know, that’s a huge, you know, airplane maker. Like you want to know that if we’re on track to renew that client, like that’s, you know, that’s critical. So you want to status up the project, the progress of that renewal compilation week to week to the executive team.
How do you do it? Super, super hard if you’re doing it manually. With Sembly, it’s a breeze. Basically, every team in their organization uses Sembly at the end of the week. They have kind of a standard. They all share their meetings into a central PMO that’s responsible for this process. So you can automatically share meetings to others on your team, to other teams, etc. so they share all their meetings to this central mini PMO. And that PMO goes into Sembly and asks Sembly to provide an end to end report from across all these teams on all their status and progress. And Sembly generates this really comprehensive document that talks about what happened this week across all the international teams for renewals. That’s magic.
Like that hasn’t been possible until Sembly came on the scene. And then actually, it’s what they do next, which is they take that and they pipe it into like an 11Labs avatar kind of thing. And they basically have a digital avatar video that speaks the report that Sembly generates, and they take that video and they send it to the C levels, and they can do this every week. That’s an insane level of up to date, like up to minute reporting for an effort that’s so distributed and so complex. So that kind of gives you a feel for what’s really possible with Sembly in your organization.
Dr. Jeremy Weisz: 24:49
Yeah, I could do a lot. What about on the marketing front? I know you have some small and large agencies, digital agencies using you.
Artem Koren: 24:58
Yeah. There’s a very large digital agency that focuses on podcast advertising out of the West Coast. That’s a customer of ours. There’s also another out of Europe that uses us. So in marketing, in some sense, similar story in the sense that all professional services agencies, companies have this flow of win the customer, deliver the work, grow the business. And so the way marketing agents use us is when they meet, so they’re usually in a pageant. It’s rare that you have someone that says, okay, I want just your agency. Usually there’s a few agencies that are being considered as vendors.
And so there’s discovery calls, there’s follow up calls, there’s pricing calls, etc. and so marketing agencies will use us in that way. They will, you know, after discovery calls they will put into Sembly, like what are the main important points for this customer to select us as their vendor? And then Sembly will tell them, and then they can say something like, you know, based on the information you have about services from prior meetings, create a customized bespoke proposal for this customer and Sembly will write something that’s laser targeted for the kind of customer that it is. The kinds of issues they’re looking to resolve, the kinds of things that are important to them in their context. You know, maybe it’s in a particular country, in a particular language, in a particular market. And so Sembly will help them do that.
Another thing that they can do is they can do like a side by side comparison. So they can ask Sembly to do research, to find online information about maybe a competitor and put it side by side, like here’s what we can do and here’s what this competitor can do and here’s, here’s how we’re better because Semblian also talks to the net and to the internet. So it can use that information as well. So those are some examples of how marketing agencies use this.
I would say, you know, we’re not. And this kind of dovetails a little bit into kind of agentic idea is that when AI is at play, the vertical specificity is still important, but starts to matter less in the technology realm. Because when you have this flow of like you’re trying to win the customer, deliver the work, grow the customer, the AI can figure out contextually the right kinds of materials to put together in your vertical because it knows about your vertical, or I should say Sembly knows about your vertical, and knows about you and your roles and everything like that. We make sure that that specific context always is a part of whatever Sembly is talking to you about.
And that’s another important difference between something like ChatGPT or Copilot or some of these other products where you have to be extremely intentional with what you specify in the context, and it can get, you know, and otherwise your mileage will vary. Like if you spend a lot of time kind of really thinking through like, what are all the details I need to include in my prompt? And then kind of like throw in a bunch of context, it will do something for you. But then next time that happens, you have to do that all over again. And sometimes you’ll hit all the context notes and sometimes you won’t. And we’ve spent a lot of time in R&D to make sure that our AI has all of the detailed background and context it needs to create a result for you that really hits the mark.
Dr. Jeremy Weisz: 28:33
It’s really interesting. I’d love to hear. I’m going to pull up. I think the pricing page here. And if we’re looking at it, obviously this could change in the future. But when people are listening to if they’re listening to seven years in the future, but we’re at Sembly.AI/pricing. I’m wondering from a feature perspective, right? Because you have probably have a lot of people making suggestions. We want this. Could you make it do this? Talk about the evolution of as a bit and, and the features that you’ve included in this product.
Artem Koren: 29:10
Yeah. So, as I do that, if, if you want, you can do kind of a slow scroll down below, below these pricing boxes, there’s actually a comprehensive side by side feature comparison. If you, if you go keep going lower, keep going, keep going. Yeah. There you go. So there’s a comprehensive feature comparison here and kind of lists all the different kinds of features that as offers. But evolution is.
Dr. Jeremy Weisz: 29:37
Because AI was a lot different when you started like in, you know, 2018, 2019. I’m sure AI was a lot different. And that’s the one thing I look at Artem when I’m choosing a, you know, software, like you have some of these fly by night companies saying it’s AI, is it AI? Is it not? Is it, you know, is it just, I don’t know if you’re seeing this, but like, I’ll see one. A couple of months ago, someone was sharing a software and then like a couple months later, it’s no longer around. Right. And so you’ve been doing this for a while and you’ve kind of seen the evolution of AI in general, right? It’s not like you just started this a year ago. So it’s interesting to see how your product has evolved with an AI has evolved a lot too.
Artem Koren: 30:30
For sure, for sure. So we are a leader in this domain of this type of product. We’re one of the founding companies in this space. So we’re one of the first, we’re not the first, but one of the very, very few, like maybe number three or so chronologically that entered this kind of space and the space being using your actual day to day work context to produce results that save you just immense amounts of time and make make your life better and make the quality of your work output better. But yeah, so certainly AI has changed a lot over these years.
I think initially, you know, the core has remained the same in that we source the context and then we try to make it productive for you.
In the first few years that meant, you know, you had searchable transcripts, you had okay quality meeting notes. I would say we had the very best meeting notes at the time pre ChatGPT, we were the first company to introduce chapter generative meeting notes by topic that didn’t exist until we did it. This was before GPT was on the scene. So there was a lot of kind of AI advancement, and we had the best task identification across the board from all the products on the market at the time as well. But the identification bit, that’s kind of where it stopped. There was very little generative stuff and the quality of the Q&A and the meeting, like if you just ask kind of just a random question about the meeting to AI, your mileage would vary. We had this feature very early, but the answers that the AI produced at the time were not so great.
And so as we went forward, AI got smarter and then the generative abilities of products got more robust. And, you know, we led the way in that space. And so we also, so we expanded in terms of what you can do with the product and also the kinds of customers you serve and how we make sure that we control privacy, security, reliability, compliance, and all those things. For in an international environment, that’s a big, big part of the product. It doesn’t matter if a company says they have AI or they don’t have AI. That’s not the point. Like what’s what does that matter to you as a user, whether there’s AI or not? What you care about is, is it safe? Does it respect your privacy? And does it give you the result that makes you go, wow, that’s what it matters. If an AI does it, great. If it’s not AI, great.
So we’re very fully on AI. We like we use a lot of bleeding edge stuff, but we do it in a very safe way. So everything we do comes through Azure and we have special agreements with Azure and how we use AI endpoints. We also have a lot of things that are local to our cloud tenants. So things don’t go outside of needlessly. So we’re it’s, you know, we have large customers on board. We have multi thousand person organizations on board. We’ve gone through very rigorous vetting process where Soc2 type two, we’re GDPR compliant, where Microsoft 365 certified, etcetera, etcetera, etcetera. So there’s a lot of things that are under the water of this iceberg to make sure that, you know, everything is there for you in a way that it should be.
But ultimately, what this allows us to do is deliver wow results in a very safe and the privacy and the privacy controlled way. And, and today, the evolution is going more and more into the kind of what are the capabilities? So for example, recently we introduced MCP access. So now if you’re working, let’s say you’re a coder, right? So like that’s, you know, a fun thing to do now. Let’s say you’re using Cursor or Claude or Codex, whatever is your go to weapon for vibe coding.
You can plug Sembly directly into your development environment with MCP. And then you can do something like this. And this is really cool. You can say in Cursor based on my discussion with the team on this meeting, spec out the feature that we discussed. Cursor will go through MCP will pull up the meeting. It will write the spec for the feature and then your next prompt is okay, implement the feature. So we’ve actually. So before people like yeah, it’s good in dev, but you still have to spec it. You still, you know, you have to spend time. No you don’t. Not anymore. You already discussed it with the team. If you discussed it across several meetings, that’s fine too. Just tell Cursor and now Cursor will,
if you’re wired up with a powerful model under the covers right with Cursor, or Claude or Codex, whatever you like. It will write a fantastical spec for you. Really comprehensive.
It will note things that maybe you didn’t even cover in those meetings. It will talk about those things. Of course, you can review the spec and tell it to include some other things, but it will be based on detailed discussions, multi-hour discussions that you had with your team. Right? When we write specs, we write based on what we think, what we remember, like, you know, we may or may not have all of the various details. You give this to Cursor. It’s got all the details. It’s going to spec the hell out of this, of this feature or this fix, and then you just get to review the spec. You actually will often be pleasantly surprised by how it thinks about it and what it uncovered that maybe you didn’t even think about. And then the next step is just say, go and do it.
So now we’ve kind of compressed the whole agentic implementation by coding thing to look at. Look at what we talked about in the meetings. Spec out what needs to be done and then do it, and then test it and then deploy it. And then tell me when it’s ready. I’m of course oversimplifying. I can already see like every advanced web coder super rolling their eyes. I get it. I use Cursor, I’m a, you know, I’m a Claude fan like all that stuff. I get it. I’m oversimplifying guys. I don’t mean this literally. Yes. We still need to sit there and watch it and help it. And when it gets stuck and goes the wrong way, yes, of course, but in an abstract way, you’re really cutting things down and bringing things very close to the source. That’s what the latest features in Sembly allow you to do.
Dr. Jeremy Weisz: 37:03
Yeah. And we could see that here in if you’re if you’re looking at the screen in the pro version, you can, MCP access is right there. There’s a bunch of other things in here. Of course, you mentioned a bunch Cursor, Claude, Codex. What are some of your other favorite AI tools that you utilize on a daily, weekly basis?
Artem Koren: 37:26
Oh man, I think so. I should mention like a lot of tools that we use now are like all partly AI in some way, like even Notion. So our knowledge management is on Notion, which by the way, of course Sembly integrates with. Notion has an AI component now. We use Attio for CRM. Attio has a fantastic AI component now. Very interesting, very powerful, not perfect. It’s got clunks, but like very, very good. And of course we integrate with Attio. So a lot of the regular tools I use now have AI component. So you know, it’s kind of hard to say like there’s AI tools and AI tools these days because all kind of everything is evolving towards some aspects of AI in some way. I would say for purely AI standpoint.
Dr. Jeremy Weisz: 38:12
So maybe you could talk tech stack too, like obviously, because that bleeds it, like you said, mentioned to you. Okay. We use this for our CRM. It uses AI. We use Notion, AI.
Artem Koren: 38:24
Yeah. Yeah. So for the tech stack, personally, I’m a Cursor guy. I started there, although there’s a lot of hoopla these days around Codex and Claude. So I might dabble there a little bit more than I have just to see kind of where they, where they gone versus Cursor. But for vibe code and like, I.
Dr. Jeremy Weisz: 38:45
Don’t chase the shiny object, right?
Artem Koren: 38:49
I can’t help it. I can’t help it. It’s a personality type. So yeah, so I like Cursor. I, I for okay. One thing that I’ve been really impressed by recently is Grok’s Imagine capability. So Grok has Imagine that can generate image and video content, and it’s the best I’ve seen in terms of simplicity, quality of outcome. It’s the best I’ve seen so far in terms of like short format video generation.
Dr. Jeremy Weisz: 39:24
I like this raccoon doing a funky dance.
Artem Koren: 39:27
Yeah, yeah, like this raccoon. But like you can basically. So what I do with it, I’ll give it like a director’s perspective. I’ll say like, you know, camera enters office, like zooms in and a worker, you know, perspective from behind, like she’s sitting at her desk, like typing on her laptop and the laptop has the screenshot on it. And I’ll give it the screenshot and you just say, go and it produces this movie for you. That’s exceptional quality. Like, I’m really impressed. Now, of course it does weird things. Of course it’s still AI, but you can, you know, you can run it a few times until you get there, but it’s really, really impressive. So Imagine is one of my favorites recently.
As far as AI tools. So I’m a Claw fan, but I wouldn’t recommend Claw for anything in production. I think for advanced users and kind of tech like techy users, like coders, Claw could be a really nice solution. It’s the most flexible of the agentic of the agentic frameworks out there. I call on an agentic framework and I distinguish between Claw and something like clawed agents, for example, because Claw is really a framework that you need to spend a lot of time tuning to your needs, but it’s extremely flexible. And if you are a tech, it’s kind of like back in the day, you know, if you like to build your own PC and you like to kind of customize your Linux, that’s OpenClaw.
Dr. Jeremy Weisz: 41:04
That’s a good analogy, I like that.
Artem Koren: 41:07
Yeah. But if you’re like a Mac person or like kind of a Windows person where you just kind of want to download the executable and double click it. You’re certainly much better looking at Anthropic offerings or OpenAI offerings today. Or my offerings for copilot.
Dr. Jeremy Weisz: 41:25
Yeah. No, that’s good. You know, it’s funny, I don’t hear many people talk about Attio CRM for some reason. I don’t know why.
Artem Koren: 41:34
It’s more of an up and comer. I think it’s along the same generational timeline as Linear, right? So I think, you know, like the precursor to Attio is something like maybe HubSpot. So it’s like Salesforce to HubSpot to Attio and then maybe from like a, from a PM standpoint, it was whatever, like, I don’t know what’s like a really old school one, like Jira to Monday to Linear. Now, I’m not like dissing on Monday or any or Jira like at all. Like, don’t take me wrong. Those are fantastic products.
But just in terms of generational, like when they came on board and what kind of like what their kind of mentality is as a product. I think there’s like a generational thing, just like, you know, we have millennials, Gen X and all that stuff. Similar is true with products. And so like Attio, I think is the same generation as Linear in terms of you can even look at their websites, like their vibe is very similar visually. Like it’s this very like minimalist.
Dr. Jeremy Weisz: 42:35
Yeah, totally.
Artem Koren: 42:36
Modern. Like, yeah. So if you look at it and then you look at Attio and Linear, like they’re pretty similar in how they think about their brand and how they think about their product. So yeah, I mean, and we like Attio so far and we like the AI stuff it brings, it’s simpler than HubSpot. It has a lot, it’s, you know, feature rich for what we’re trying to do. So in that sense, it’s very good. I also think like these generations kind of reflective of the philosophy of product that’s shifting, you know, with the advent of AI, like I think products themselves are changing. And that’s something I wanted to make sure we kind of touch on this podcast, which is this idea that, you know, classical products were about features and applications themselves were kind of organized data stores, and they were really tuned to present certain vertical data in a certain way.
So like your CRM is like a very organized data store for sales data. Your project management system is a very organized data store for project data. But, and so it’s great for what it is, right? Like it makes, it enables a lot of new productive outcomes as a result of having that structured data and the result of kind of the features that it includes, but it’s ultimately up to you as the user to use that organization and the availability of that data to achieve some kind of goal or results. And that’s the model of all classical software until very recently. Some software still, you know, still is the main focus. But we’re in a generation of products now that are enabled agentically.
What does that mean agentically? If again, I’m going to hyper simplify and I’m just going to say that agentically is something that uses AI to do multi-step things and make decisions along the way, and includes abilities to use tools along the way. So instead of just being like one off Q&A, like ChatGPT, it’s something that can like you give it a goal, like, you know, make this software better and it can reason about that goal. It can figure out what steps it needs to take to reach that goal. And it can make use of tools like coding tools or whatever to get it to get itself to that goal. That’s what I mean by agentic. So it’s multi-step reasoning has some tools.
And so that idea when you productize it, when you wrap some kind of product around it. What you’re getting is, rather than this verticalized feature oriented product, you’re getting a goal accomplishment product. So a CRM classically was something that gave you really useful features to do your sales, right? Awesome. It’s super useful. Like can’t live without it. But like next gen CRM, like agentically empowered CRM is something that doesn’t get you the features to do it right, but it’s something that can collaborate with you to achieve your sales goals, whatever that may mean. So it really levels up the level of abstraction for the kinds of other outcomes that it can provide.
And also what that means is that very often in these products, you can now start to span outside of your immediate zone of responsibility because to generate a great proposal, you may need to interact with the design team. You may need to interact with, you know, the support team. Like were there any issues with this customer? The product can do that. It can go walk around, it can go figure out, it can pull it all together and it can give you the goal. It’s a little bit of like a wood filler wish. Fulfiller. And so modern, like next gen products, ourselves included, are more in that category. And we’re, we’re putting a lot of R&D into those aspects of product that, you know, in the next few months will come out that are really going to, you know, knock the pants out of people.
Dr. Jeremy Weisz: 46:58
Artem, first of all, thank you. Thanks for sharing the journey and just your thoughts on AI. You’ve been doing this a while. I don’t know if you, I had one question. I don’t know if you have a minute. We’re right at the top of the hour if you have to go. No worries. But I just, my last question was just I was going to ask about we don’t get here alone. Right. And some of the mentors, colleagues have most influenced you. And since we only have maybe a minute or two since I don’t want to make you late, maybe I’ll just start, talk about Gil and some of the best advice Gil’s given you or and how he’s helped in the journey.
Artem Koren: 47:38
For sure. So Gil and I. So Gil is our co-founder, Gil Makleff. The CEO of Sembly. I have a relationship with Gil that goes back close, getting close to a quarter century. So we worked a lot together. Gil’s background is deep in consulting and professional services, and so he had an exit to Microsoft and then later on, exit to Ernst and Young with a consulting company. And I worked with him a lot during kind of for a better part of a decade. So Gil, we’re, we’re great balancing forces because I’m really the shiny object guy. And Gil has this kind of great intuition for kind of when to hedge and when to kind of slow down and, you know, when to make moves. And that’s, you know, that’s something that I’ve kind of started to absorb over time.
We’ve had a very tumultuous seven years as a humanity. You know, when we started, right away came Covid. That changed a lot of things. Then came the, you know, that then came a war in Ukraine that changed a lot of things. Then came the AI revolution that changed a lot of things. And so with every one of these upheavals, there was kind of managing that needed to happen. And Gil is very good at that, kind of looking at that high level and that longer term horizon.
My focus tends to be, you know, a lot more kind of like what’s the vibe today. And, you know, how can we affect what’s going on today? And it’s interesting because as the world turns, those different forces become more and less important in the situation that we’re in. So sometimes it’s important to like, just go, go, go, like adopt the latest, like create the latest, put it out there. And then there’s other times when it’s like, well, okay, step back. Like we need to catch up. We need to raise, we need to capital up. We need to hedge because, you know, things are coming down. So I think that’s been a great partnership. And you know, I’ve learned a lot from that and continue to.
Dr. Jeremy Weisz: 50:08
Thank you Artem. Everyone check out Sembly.AI. Check out more episodes of the podcast and we will see everyone next time. Artem, thanks so much.
Artem Koren: 50:17
Jeremy what can I, can I throw out one thing for you?
Dr. Jeremy Weisz: 50:19
Go ahead, go ahead.
Artem Koren: 50:21
So for listeners of this podcast, if you register for a free trial with as, which I encourage you to do, it’s noncommittal. You’ll get a week for free completely. But if you use the coupon code Podcast2026, you’ll get a discount. So I encourage you to try.
Dr. Jeremy Weisz: 50:38
Podcast2026. Check it out. Awesome. Thanks, everyone. Thanks, Artem.
Artem Koren: 50:43
Thank you.
