Smarter Signals, Stronger Agents, Better GTM

04:29:35.040 –> 04:29:43.139
Julia Nimchinski: What a pleasure! And we are about to see a demo explorium.

1487
04:29:46.510 –> 04:29:47.760
Julia Nimchinski: Amazing.

1488
04:29:48.420 –> 04:29:50.320
Julia Nimchinski: Welcome back, Omar.

1489
04:29:50.930 –> 04:29:52.890
Julia Nimchinski: Smarter signals.

1490
04:29:53.600 –> 04:30:03.120
Julia Nimchinski: better agents and Gtm. Welcome to the show again. And how would you like to set the stage? Shall we just dive into it?

1491
04:30:03.590 –> 04:30:08.640
Omer Har: Sure I can. I can just run the demo. I don’t. It’s not a problem to share right? Share my screen.

1492
04:30:08.640 –> 04:30:09.419
Julia Nimchinski: Yeah, sure.

1493
04:30:09.630 –> 04:30:10.230
Omer Har: Sure

1494
04:30:23.210 –> 04:30:26.399
Omer Har: I’m waiting for you for you to kick me off, so just.

1495
04:30:28.630 –> 04:30:31.080
Julia Nimchinski: Oh, are you waiting?

1496
04:30:31.080 –> 04:30:34.430
Omer Har: I’m waiting for you to just tell me that it’s okay. Can I stop? Then.

1497
04:30:34.430 –> 04:30:35.830
Julia Nimchinski: Oh, yeah. Yeah. Yeah.

1498
04:30:35.830 –> 04:30:49.380
Omer Har: That’s completely fine. So so Hi, everyone, I’m I’m Omi, and I’m the co-founder and CEO of explorium. I’m gonna in the next 1015 min. I’m gonna show you a bit about our product

1499
04:30:49.570 –> 04:31:10.969
Omer Har: and kind of tell you more about what we do. Exploring is a data aggregator. We aggregate data coming from about 50 different sources, and we create one or harmonize the data in order to create one layer. We call that fabric sometimes. And the idea is that that fabric allow you to interact with our data anywhere, from any need, from company, from a prospect

1500
04:31:10.970 –> 04:31:30.879
Omer Har: to an account, to a location, and you can move away. You can move from 1.1 entity to the next in a very easy way that you can. Then you can then learn and use the data. The way that we work is by providing what we call an Api suite. Of all of the endpoint we have more than 25 endpoints that allow you to discover anything that you want

1501
04:31:30.880 –> 04:31:37.000
Omer Har: on a company, or a person, or even location. Anything that you want in order to help you get to.

1502
04:31:37.320 –> 04:32:06.879
Omer Har: you know, to create, go to market motion. We work with many companies in the area, from from better data and better sophisticated, more sophisticated agent in order to you know, solve your go to market problems. So let me show you kind of a couple of the things that we’re doing. And specifically, I’m going to show today kind of the our Mcp. And I will explain what that means in a sec. Let me just share.

1503
04:32:15.400 –> 04:32:39.800
Omer Har: So this is a demo app that we created specifically to show how powerful it is to connect our Mcp with specifically in this particular K. Claude. But you can use any. Llm. Specifically, we were with this product yesterday in product hunt, and we won the 1st place by a landslide specifically, because I think that once you start looking at, that’s very powerful.

1504
04:32:39.890 –> 04:32:53.159
Omer Har: Mcp is basically a technology. It’s maybe a big acronym. But it’s a simple technology that allow an Llm. Any agent think about Gpt or Claude. Anyone that you’re using in

1505
04:32:53.160 –> 04:33:23.120
Omer Har: can connect to basically tools. So we took all of our 25 endpoints. It’s a lot, and we connected that through an Mcp. And allowed the Llm. To actually learn something about our Apis and start running with what I’m showing you here is. You know, there’s no magic here. It’s our Mcp. With vanilla clawed within a very simple ui. The whole things took us a couple of day to build. We didn’t add any special, any special magic. It’s just our Mcp, our tool connected to it, to A

1506
04:33:23.750 –> 04:33:31.369
Omer Har: to connect it to an Llm. So let’s start with something very simple like, let’s say you want to know whether

1507
04:33:32.470 –> 04:33:41.619
Omer Har: Does adobe is using Apache zookeeper.

1508
04:33:42.849 –> 04:33:46.200
Omer Har: So Apache zookeeper is a type of

1509
04:33:46.279 –> 04:34:11.190
Omer Har: of open source system that allow you to send requests and basically queuing system and monitoring system as well. So what we’re seeing here is that we send it down to Claude. Claude. Now have all of our Mcp, all of our endpoint to kind of use. And now he’s trying to actually solve the problem and and understand does adobe is using Apache zookeeper the 1st thing that he’s going to do. And again, I didn’t do it. I didn’t tell him to do anything

1510
04:34:11.189 –> 04:34:18.890
Omer Har: special here. There’s no special thing is just a Claude connected to an Llm. You can do the same thing using your Claude desktop.

1511
04:34:19.040 –> 04:34:45.879
Omer Har: The 1st thing that he’s doing. You need to understand what adobe is right? So you need to match that business towards some id. The reason that it’s so important that if you’re using just a simple kind of just going into the web and looking for adobe, it might work because adobe is a huge company. However, there is a company named Amazonia, which is a supply chain company. And if you search Amazonia in Google, 99% of the cases, you won’t find it, because Amazon will be there to grab your attention.

1512
04:34:45.880 –> 04:35:03.290
Omer Har: So from that perspective, you know that that’s kind of one of the things so adobe is easy as Amazonia is hard. You need to 1st match it. And this is exactly what you’re seeing. And basically what is happening in the background is that the agent is talking to our system in order to understand like which one which adobe is it? And what is the Id.

1513
04:35:03.380 –> 04:35:18.370
Omer Har: Once that is good, then it can actually use a different endpoint in this particular case is business technographic. In order to understand what is the technology that this business have? And you can see there is a long list of different technologies that adobe uses just to kind of

1514
04:35:18.470 –> 04:35:26.830
Omer Har: making sure that you kind of just understand a bit more. Most of this data is actually collected from job description. So when adobe

1515
04:35:26.830 –> 04:35:51.329
Omer Har: publish a job description in many cases they will ask for we want experience in, let’s say, Amazon, Ec. 2. Or in this particular case, adobe is using Apache zookeeper. So if they in some point we identify that adobe will look look for experience in Apache zookeeper, probably more than one time, and we pull that in and say, well, so that means that adobe is actually

1516
04:35:51.935 –> 04:36:02.034
Omer Har: you know, using that internally, otherwise, there’s no reason for that. That’s just an example. Another point that we can actually look at is what if we want to know more about

1517
04:36:02.520 –> 04:36:11.260
Omer Har: When we want to know more about the company, let’s say a different company. We want to understand more about what is the strategic changes that this particular company have

1518
04:36:11.290 –> 04:36:35.269
Omer Har: so kind of one of the things that many people are currently doing is actually using the web and kind of web research agent in order to understand, in order to actually read more information about the company and learn more. But there’s a lot of data that is not available straight up in the web, for example, snapshot of the website. So if you think about, if you want to learn what happened specifically from a positioning perspective, for example.

1519
04:36:35.380 –> 04:36:47.679
Omer Har: for a specific company. You can look at the website in 2 different snapshots, let’s say, 3 months apart and analyze the result in order to understand what is different there. Okay, so let’s say, get the web changes

1520
04:36:49.779 –> 04:36:50.940
Omer Har: for up to you.

1521
04:36:52.380 –> 04:37:17.170
Omer Har: So again, what will happen here is, instead of trying to do that on the web, where not necessarily, you’ll get this type of information. You can then go into one of our endpoints and actually query for that information in order to understand how we can change that in the same way that we did that for adobe, when the 1st thing that we need to do is the entity matching, and only then start running the rest. Then we’ll need to do that as well. Here 1st is going to match around

1522
04:37:17.169 –> 04:37:25.950
Omer Har: the business that should take just to take a couple of seconds, and once he identify the business, he can then try to look for web changes specifically for Atio.

1523
04:37:27.150 –> 04:37:31.570
Omer Har: Okay, that’s kind of how that’s kind of how it works. And you’ll probably see that in a second or not.

1524
04:37:32.070 –> 04:37:54.269
Omer Har: When you think about the kind of a bigger kind of what is the plan for? What is the plan for Mcp overall? Here you go now. He’s running for that. If you think about how Mcp can work in a more general way. So we believe that. You know, if you think about Gpt, for example, Gpt. 0 3 went live about 3 weeks ago, as part of kind of the.

1525
04:37:54.480 –> 04:38:16.699
Omer Har: you know, release. They actually let him run an IQ test the same thing that many people running themselves. He got 130, which is 2 standard deviation more than the the average, which means that 96%. According to that test, specifically, 96% of the human population around the world is actually less smart than O, 3.

1526
04:38:16.700 –> 04:38:40.480
Omer Har: Which means that what we want to say is saying, okay, so the agent can be very smart if we give him all of the information. If we give him an Mcp with all of the information possible, what can. He does? What can he do? And I think that’s kind of how we think about, how can we think about our tool set? You’ll take explorium is a tool set that you plug in into your any Llm. And this Llm. Will become a go to market. Intelligent machine.

1527
04:38:40.480 –> 04:38:53.920
Omer Har: In this particular case you can see that when we look for Atio, which is a type of a kind of a new age. Crm, you can see that they have many changes in the website. They change from new object to the next Gen. Of Crm. Because they released different features.

1528
04:38:53.919 –> 04:39:18.620
Omer Har: We know that there is a security focus because there was no mention of security certification 3 months ago, but now the Gdpr. Ccpa. And other certificate, and you can see the different. They didn’t have any customer segment. Now they do so. All of those things is not available on the web. However, if you give an agent this freedom of not only testing the web, but also going into every one of those b 2 b databases that we

1529
04:39:18.619 –> 04:39:21.940
Omer Har: hold. As I said, more than 25 endpoints.

1530
04:39:22.279 –> 04:39:38.719
Omer Har: You have more things or more information that you can pull, which is super interesting, especially if you case in a case where you need that data. For, let’s say, prospecting research, and so on. If we’ll take another example which I which I love is get the email

1531
04:39:39.430 –> 04:39:41.520
Omer Har: or visible

1532
04:39:46.465 –> 04:39:46.970
Omer Har: Ciso.

1533
04:39:47.760 –> 04:40:02.160
Omer Har: So the reason that I love this is because this is something that many, many go to market practitioners is trying to do. They want to reach out to the Ciso invisible. Let’s say I’m trying to sell some in a security system, and they’re trying to

1534
04:40:02.380 –> 04:40:26.140
Omer Har: use. And we’re trying to sell directly to visible. And they want to get to the person of their Ciso. The reason that I’m like this this example is because there isn’t any Ciso for visible. They don’t have a Ciso. However, they do have a director of information security. So now the system will need, or the agent will need to automatically figure that out.

1535
04:40:26.240 –> 04:40:45.400
Omer Har: find this person and use that person and find the email for that person. It can take a bit of a couple of minutes to run that. But the main point will be that the agent again as and I didn’t do anything special for this particular agent, or the only thing that I gave him is and I refer

1536
04:40:45.580 –> 04:40:50.249
Omer Har: to the agent as him. For some reason the only thing that I did it just gave him

1537
04:40:50.270 –> 04:41:11.839
Omer Har: the our Mcp. So once I connected it to the Mcp. He started working and kind of looking for different ways that he can then solve the question that I ask and spin this particular case to find the Cecil. When we think about the future, we think about what we call usually autonomous future, where the agent will be will have the autonomous, the autonomousity.

1538
04:41:11.840 –> 04:41:30.569
Omer Har: I guess, to actually choose what to do and how to solve a problem. And what exprogram does is provide him with the necessary tools in order to solve, go to market problems. We think about that as a shift, because up until now Mcps usually mean mainly for integration. We see that as a product, Mcp as a product.

1539
04:41:30.570 –> 04:41:45.529
Omer Har: If you release. If you take your Mcp and you connect it, you take our Mcp and you connect it to an Llm. The next step that will happen is that this Llm. Will become a prospecting machine, a go to market, intelligent machine that you’ll be able to use. That’s kind of how we think about

1540
04:41:45.530 –> 04:42:00.660
Omer Har: about it. Most of our customers, I said in the beginning, is customer, that building go to market intelligence system on top on top of ours and on top of our infrastructure and data. It can be, you know, a qualification agent. It can be

1541
04:42:01.451 –> 04:42:04.820
Omer Har: outreach agent like to send actually emails.

1542
04:42:05.080 –> 04:42:16.260
Omer Har: it can be around just providing the base layer on top of which you’re going to do research like we’re showing here is here. And this is kind of usually how we we work with those teams.

1543
04:42:17.440 –> 04:42:44.609
Omer Har: just to show you just to notice, like, just to show you kind of what he’s doing here. Not only he matched the business. That’s the easy part. Now. He’s trying to look for Ciso. He couldn’t find that I don’t see a specific Ciso title match. So now he’s trying to other, let’s say, creative ways. For example, look for query security in order to figure out what he can learn and what he needs to do. And after all of this he’ll very soon will get to a point where he said, Okay, I know who I need to reach out to

1544
04:42:44.610 –> 04:42:51.580
Omer Har: that will be a director of of it. I know that we’re very short in time, so I I want to be quick with that.

1545
04:42:54.860 –> 04:43:02.560
Julia Nimchinski: Where one of the questions that we’re receiving right away is, how long does it? How fast can actually teams get this live?

1546
04:43:03.050 –> 04:43:06.000
Julia Nimchinski: And how long does the training take, or how does it work.

1547
04:43:06.520 –> 04:43:29.439
Omer Har: So there’s no training involved here. The only thing that you need to do is to take your Api key for Cloud, the Api key that you get from explorium. Combine them together and kind of work together. With that, you can do that with Langgraph Langgraph automatically support Mcp. So if you’re working with Langgraph or Glangchain. That should be a breeze.

1548
04:43:29.440 –> 04:43:39.810
Omer Har: If you want to just try it out, you can. You, if you want to actually hook it up to a cloud and try it out yourself. You can do that with cloud desktop or in our demo up, just to show that around

1549
04:43:41.320 –> 04:43:51.680
Omer Har: if you are implementing something, if you want to implement something like that, just a chat bot that can give you information and kind of insight on on specific

1550
04:43:51.950 –> 04:44:06.310
Omer Har: customers just for you. So imagine that you have a product, and within within it you have a chatbot similar to what you have here that allow you to do that. Then, in particular case, all you need to do is basically. And this is what we did here connected that to land graph.

1551
04:44:09.160 –> 04:44:10.240
Omer Har: framework.

1552
04:44:12.520 –> 04:44:21.450
Julia Nimchinski: Very cool. And one other question here. Truly, a lot of them. How do you compare? How do you compare it to other players in the space.

1553
04:44:21.810 –> 04:44:24.580
Julia Nimchinski: And who are you even competing with.

1554
04:44:25.140 –> 04:44:40.354
Omer Har: Cool. That’s a good question. So traditionally, we, we are competing with data provider like us that provide kind of layers of data. So you can think about Apollo. I/O, as an example, is an example of competitors. There is other, obviously,

1555
04:44:40.700 –> 04:45:03.750
Omer Har: usually what those guys in most cases, what you’re seeing. Those guys are actually focusing on creating their own agents like Apollo as an example. What we’re focusing more is by providing this layer that agent can connect with us. We believe that the real technology, or from our perspective, what we know how to do best is actually to power that layer of data that is integrated

1556
04:45:03.750 –> 04:45:19.889
Omer Har: and efficient from a from a harmonization perspective is efficient. So you can move from one entity to the next and do that pretty quickly, and then use that in order to to solve, go to market question, whether to find the email for

1557
04:45:19.890 –> 04:45:29.380
Omer Har: Baraki goes, which is the the one that security that that focus on security for for visible or other, you know, go to market issue that you have.

1558
04:45:31.460 –> 04:45:41.949
Julia Nimchinski: Very helpful. And, Omart, can you share more? Use cases? Probably some of the top ones that you’re seeing, and we’d love to hear more about your latest partnership with outreach.

1559
04:45:42.530 –> 04:46:00.569
Omer Har: Sure. So we work a lot in the agentic, you know, with agent company, that build agent, and one of the kind of the earliest partnership that we did was actually with outreach I owe. We’re working with them with over a year now. I think that the team is outreach is really kind of is

1560
04:46:00.610 –> 04:46:13.809
Omer Har: future forward. When you think about kind of what they’re trying to do, they’re building, you know, things that up until 2 years ago was imaginary to most go-to-market tools and kind of what they want is to build a completely new.

1561
04:46:14.270 –> 04:46:27.230
Omer Har: a way to interact with how sales and Aes interact with customers, and how efficient you can be to do that. What we do is provide all of the data layer and infrastructure that they need

1562
04:46:27.230 –> 04:46:53.800
Omer Har: in order to run them their their agent. We’re already in Beta. We. I think the Ga. Is in a couple of months away, but the idea will be that every you know, every customers any customer from outreach can come into come to outreach, get a a an agent that will help him prospect and then send emails to. And you can do that on our data and get the best email best personalization based on their technology.

1563
04:46:56.720 –> 04:47:03.949
Julia Nimchinski: Super helpful. And people are asking about the Roi, and how fast can you actually see it.

1564
04:47:05.050 –> 04:47:19.720
Omer Har: Yeah, that’s a good question. So we can. You know, we work with also a lot of smaller startups that you know, they move very quickly, and they need data in order to do multiple things. So one of them, for example, is find you

1565
04:47:19.860 –> 04:47:24.110
Omer Har: find new prospects that you want to reach out to. It’s

1566
04:47:24.370 –> 04:47:42.259
Omer Har: as as it mostly based on Api. And we have a lot of experience kind of working with companies that needs that type of of information. We can do that in days. So once you connect to us, it can take days until you’re you’re up and running with a valid solution that you can then sell to the market.

1567
04:47:43.460 –> 04:47:50.149
Julia Nimchinski: Amazing, Omar, what would be the best next logical step? And how can people find more about explorian.

1568
04:47:50.810 –> 04:48:18.419
Omer Har: Cool. So I think that our website is the 1st thing that I’m going to do what I’ll also suggest within the website. If you register, you can go into our Mcp playground to play around exactly how I did with with the app here. I think that would be the best. Usually what happened is that people play around with the playground and reaching out directly after kind of asking more questions. So I’ll be happy to to do that, and you can reach out to me. Of course I’d love to have a chat.

1569
04:48:19.250 –> 04:48:20.600
Julia Nimchinski: Thank you so much again.

1570
04:48:20.600 –> 04:48:21.060
Omer Har: Thank you.

1571
04:48:21.060 –> 04:48:22.109
Julia Nimchinski: Pleasure hosting you.

1572
04:48:22.110 –> 04:48:23.520
Omer Har: Thank you. Bye-bye.