The Largest Agentic CS Deployment in B2B

05:28:22.690 –> 05:28:25.080
Julia Nimchinski: Welcome, Alok Shukla!

1749
05:28:25.320 –> 05:28:31.589
Julia Nimchinski: This is going to be really interesting one, the largest agentic Cs deployment in B, 2 B.

1750
05:28:32.680 –> 05:28:35.040
Julia Nimchinski: Welcome to the show, Locke. How are you doing.

1751
05:28:36.070 –> 05:28:38.059
Alok Shukla: Pretty good. Can you hear me.

1752
05:28:38.410 –> 05:28:39.190
Julia Nimchinski: Yeah.

1753
05:28:40.460 –> 05:28:41.570
Alok Shukla: Absolutely.

1754
05:28:41.730 –> 05:28:43.889
Alok Shukla: I was excited to join you guys.

1755
05:28:45.250 –> 05:28:54.670
Julia Nimchinski: Amazing. I know you had a special plan for this one. How do you want to take it? Do you want to do a presentation. A demo. What flow are we in.

1756
05:28:55.400 –> 05:29:06.009
Alok Shukla: Yeah, we’ll do with something, both mostly to start with a quick couple of slides, just to kind of set up the context. But after that it would be mostly live product, demo

1757
05:29:06.150 –> 05:29:13.139
Alok Shukla: and hope to complete and around 1520 min. And then we can go for open ended question answers.

1758
05:29:13.630 –> 05:29:14.420
Julia Nimchinski: Awesome.

1759
05:29:15.400 –> 05:29:16.480
Alok Shukla: Sounds good.

1760
05:29:17.120 –> 05:29:22.839
Alok Shukla: Okay, let me share my screen.

1761
05:29:33.220 –> 05:29:35.410
Alok Shukla: Okay? So

1762
05:29:36.030 –> 05:29:43.620
Alok Shukla: I’m alok, and I’m the founder of funnel story. And thanks for hosting me for this agentic AI Conference.

1763
05:29:43.970 –> 05:29:44.815
Alok Shukla: So

1764
05:29:46.220 –> 05:30:14.100
Alok Shukla: product story is, almost a 3 year old startup that has focused on customer support intelligence primarily focused use case are for post sales, the renewal teams, the customer success teams, and we happen to have developed for almost last last one and a half years one of the largest agentic orchestration that any of our competitors do in this part of the if part of the vertical that we deal with.

1765
05:30:14.420 –> 05:30:19.630
Alok Shukla: So that’s why I would love. I’m very excited to talk about today. And let’s get on with that.

1766
05:30:20.840 –> 05:30:33.489
Alok Shukla: Okay? So for a story, as I said, AI, native customer support Intelligence Platform entirely builds on agents, and the way we think about agents is a lot different

1767
05:30:34.080 –> 05:30:41.330
Alok Shukla: compared to how many of the companies that have bolted on agents looks like

1768
05:30:41.470 –> 05:30:47.873
Alok Shukla: so unlike having a product where you kind of trust trap on an agent, or you put

1769
05:30:48.785 –> 05:31:17.699
Alok Shukla: do a kind of AI washing on top of that funnel story was created as an agentic platform to do a lot of its work, how the products operates from inside, how the products operates some of the smaller things it does, how it executes its features, it functions, how it interacts with the customers, customer success teams. The agentic architecture is built into everything. So when we say that you are dealing with agentic swamp.

1770
05:31:18.730 –> 05:31:23.789
Alok Shukla: what it means is that any place you test the product it built off agents.

1771
05:31:24.570 –> 05:31:30.290
Alok Shukla: So we will talk more about that. But if there are any questions I would be happy to take them as we go along.

1772
05:31:32.020 –> 05:31:32.610
Alok Shukla: Okay.

1773
05:31:33.380 –> 05:31:39.809
Alok Shukla: 1st of all, before we even go into the product. A good thing is to kind of talk about

1774
05:31:40.100 –> 05:31:47.279
Alok Shukla: what does even this agentic song even achieves like if you are with funnel story, what it will be able to achieve.

1775
05:31:47.450 –> 05:31:52.850
Alok Shukla: So there are many things funnel story can do. But broadly, there are

1776
05:31:53.370 –> 05:31:57.090
Alok Shukla: key issues that we attempt to solve. Number one.

1777
05:31:57.210 –> 05:32:09.819
Alok Shukla: the one of the single biggest problems in adoption of any kind of revenue or customer success platform as it stands today that they take months to deploy, if not weeks.

1778
05:32:10.890 –> 05:32:16.840
Alok Shukla: and final story has literally reduced that time to an hour or a couple of hours.

1779
05:32:16.960 –> 05:32:37.920
Alok Shukla: and that is log, primarily due to the agentic automation which ensures that it does not require any engineering work. It does not require any professional service. It just builds the entire stack without asking the questions, without requiring to manually have manual intervention. It just does things as the agents are supposed to do.

1780
05:32:38.840 –> 05:33:00.130
Alok Shukla: Then funnel story is has another agent which basically, we call this the prediction agent. It calculates the the churn risk or the expansion potential of the accounts automatically every day, analyzing thousands of accounts and generating predictions. That’s done entirely, autonomously.

1781
05:33:00.410 –> 05:33:04.230
Alok Shukla: And it calculates the very unique thing. And we’ll talk in the demo

1782
05:33:04.660 –> 05:33:12.240
Alok Shukla: number 3 is that and this is, we are very should be proud of this, and we accidentally pivoted to this feature.

1783
05:33:12.500 –> 05:33:25.330
Alok Shukla: There are a lot of tools in the industry, for example, which are very vertical. AI, you have AI zoom, you have AI on Salesforce, AI on Hubspot or AI on many other tools.

1784
05:33:25.590 –> 05:33:33.019
Alok Shukla: The problem that industry faces is that a lot of, especially for large enterprise faces that they have

1785
05:33:33.190 –> 05:33:35.800
Alok Shukla: a swarm of enterprise data. They have.

1786
05:33:36.674 –> 05:33:43.620
Alok Shukla: unstructured data and structured data living in Crm support tickets, product usage bugs

1787
05:33:44.210 –> 05:33:49.049
Alok Shukla: your conversation, your emails, your chats, your notes, you name it.

1788
05:33:49.400 –> 05:34:08.269
Alok Shukla: And they are not looking for siloed intelligence siloed ais. They are looking for the entire data to work together so that they can search and reason. Part of the search problem, especially for text, was being solved by companies like Gleam Covio. But what funnel story has done is goes way beyond that.

1789
05:34:08.280 –> 05:34:21.449
Alok Shukla: Where we have introduced the reasoning capabilities with this agents. So one of story becomes the 1st platform in the industry where you can search and reason your entire customer data to a single platform. And I’m extremely proud to show that today.

1790
05:34:22.040 –> 05:34:36.220
Alok Shukla: then we have more things which are basically about the usual automation, autonomous nature of the tasks that are done by customer success people today but can would be automatically done by AI, and we will show you how they are done.

1791
05:34:36.360 –> 05:34:41.719
Alok Shukla: So that’s kind of a very quick view of what funnel story is capable of doing.

1792
05:34:42.336 –> 05:34:48.010
Alok Shukla: But and we will touch many of these points. Go one by one before I

1793
05:34:48.240 –> 05:34:50.550
Alok Shukla: kind of jump into the product. That is a

1794
05:34:51.090 –> 05:35:14.059
Alok Shukla: a conceptual understanding of the product concept that functionally connects with your structured and unstructured data. Something I said, and we have a Felix, or what we call our agentic AI. We have given a name so that it’s simple to understand. It basically connects with your entire structured and unstructured data to build a hyper, dense knowledge graph

1795
05:35:14.510 –> 05:35:26.079
Alok Shukla: which is ultimately responsible for all kind of interaction, whether you are out of the box feature, or you are interacting to do AI search. And AI reasoning on top of that.

1796
05:35:26.500 –> 05:35:34.470
Alok Shukla: Okay? So at this point of time, I want to kind of start talking about the product, and I will come back to slides if I need to.

1797
05:35:34.640 –> 05:35:39.140
Alok Shukla: So let’s start with the 1st thing the funnel story

1798
05:35:39.530 –> 05:35:45.780
Alok Shukla: is can connect with your structure and our structure data. And I want to kind of start with how that does that happen?

1799
05:35:47.360 –> 05:35:52.170
Alok Shukla: So far, story can connect to all kind of sources of data.

1800
05:35:52.300 –> 05:36:01.730
Alok Shukla: whether it’s your your new style, database, your sequel, database or new SQL, database, or all kind of data simple database inside the house, or data warehouses

1801
05:36:02.130 –> 05:36:06.929
Alok Shukla: and host of application, and we keep on adding these application almost every day.

1802
05:36:07.800 –> 05:36:20.379
Alok Shukla: But the interesting part of follow story is that from an agent perspective, it does not actually take care too much about where which application you are connecting. It cares about something else.

1803
05:36:20.670 –> 05:36:25.159
Alok Shukla: It cares about that it understands your data intuitively.

1804
05:36:25.770 –> 05:36:34.399
Alok Shukla: it understands support ticket data, it understand conversation data. So, for example, if you offer, if you show from a story an agent that, hey? This is an app.

1805
05:36:34.690 –> 05:36:50.159
Alok Shukla: It will automatically start to figure out, okay, this looks like a conversation data. And these are the apps. For example, I know where conversation data can be available, and I will fetch that data automatically. You don’t have to tell me. Beyond that I will inquire. I will understand. And I will fetch the data back.

1806
05:36:50.600 –> 05:36:57.759
Alok Shukla: And that’s basically takes away multiple level of human intervention which is costly and which is time consuming

1807
05:36:58.050 –> 05:37:04.959
Alok Shukla: the moment you connect the data. For example, funnel story will automatically build out this entire configuration. In minutes

1808
05:37:05.390 –> 05:37:08.810
Alok Shukla: it will figure out what data, for example, I want.

1809
05:37:08.970 –> 05:37:16.850
Alok Shukla: build queries in some of the cases of our data warehouse what the data looks like, validate it. And just ask you what you need to done

1810
05:37:17.190 –> 05:37:26.470
Alok Shukla: so entire data, configuration and maintenance done by the agent in a matter of minutes, something that takes months

1811
05:37:26.850 –> 05:37:28.429
Alok Shukla: to get to this point

1812
05:37:28.690 –> 05:37:35.219
Alok Shukla: from a just efficiency and productivity release perspective. Not only it saves money, it saves something more valuable, which is time.

1813
05:37:36.120 –> 05:37:37.570
Alok Shukla: So we got the data.

1814
05:37:37.910 –> 05:37:45.110
Alok Shukla: Now, what we do, the second thing, what happens in a normal customer data process is that

1815
05:37:45.650 –> 05:37:58.729
Alok Shukla: you? You will deploy a tool and then you will expect some human or somebody who has a deep knowledge of data to explain the data so that the traditional Saas tools can make sense of it. Generate some dashboard, some data from tooling.

1816
05:37:58.850 –> 05:38:10.309
Alok Shukla: But with agent. AI. You don’t have to do that goes to a next step. When you have this data, it starts to undersco back in time and create a hyper, dense graph

1817
05:38:10.540 –> 05:38:13.950
Alok Shukla: to understand. For example, let’s say you collected 10 years of data

1818
05:38:14.310 –> 05:38:35.469
Alok Shukla: customer connected to your product customer was chatting in in your slack, or it was that you were sending an email or filing a support ticket, whatever they were doing across all types of customer, all types of users. But a story would go back in time and create a hyper, dense graph of understanding all the pathways your customers ever interacted.

1819
05:38:36.930 –> 05:38:45.519
Alok Shukla: There is no tool that does it before we filed a patent for this, and this is done. And, by the way, it’s not done by any trigger, it is done autonomously.

1820
05:38:45.740 –> 05:38:55.829
Alok Shukla: It goes beyond it, and starts to understand how your customers get value. Who are your best customers, where your customer drop off, which has the most interesting interaction point.

1821
05:38:56.050 –> 05:39:03.159
Alok Shukla: So the level of intelligence it generates before even the agent talk starts. Talking to you

1822
05:39:03.450 –> 05:39:09.480
Alok Shukla: is something that you expect. You expect somebody to know more about your behavior by themselves

1823
05:39:09.620 –> 05:39:15.849
Alok Shukla: without us spending time just to kind of get extract intelligence out of you. That’s power of aging here.

1824
05:39:16.450 –> 05:39:25.139
Alok Shukla: So what happens is in that in the 1st R itself, data is in data has been hypertensed, analyzed conversation have been analyzed.

1825
05:39:25.420 –> 05:39:33.740
Alok Shukla: We now understand when people chat to you what you were doing before what you did before. File the support ticket. What you did after how much time you take, how many people dropped off?

1826
05:39:33.870 –> 05:39:39.849
Alok Shukla: Pretty much everything right? And then, for a story does. Another thing

1827
05:39:40.020 –> 05:39:54.839
Alok Shukla: is that it actually has identified the best pathways for your customers to succeed. So if you connect click on this button, it will automatically generate an adoption funnel based on your historically best data. Best pathways your customers have done.

1828
05:39:55.430 –> 05:40:02.399
Alok Shukla: and it will generate a pathway which you can just use to drive the new customers for the same adoption journey.

1829
05:40:04.330 –> 05:40:17.510
Alok Shukla: Why, that is important is because it requires months of consulting work by the companies to build this kind of work, which is done by 9 people in a committee room doing.

1830
05:40:17.640 –> 05:40:28.200
Alok Shukla: arguing with each other what is the best journey, and here the AI and the combination agent AI is building that for you, just serving of you in the 1st one and a half hour itself.

1831
05:40:28.470 –> 05:40:30.970
Alok Shukla: Huge savings again, money and productivity.

1832
05:40:31.470 –> 05:40:35.280
Alok Shukla: So we have done that now what we can do next.

1833
05:40:35.540 –> 05:40:38.990
Alok Shukla: So the 3rd thing now financially does it.

1834
05:40:39.160 –> 05:41:06.800
Alok Shukla: that because it understands your usage, your conversations, your meetings, your chat, your everything. It starts to do what is very normally called in the health scoring it. Basically, we hold a patent, by the way, for the only company which combined usage, conversation, external business, intelligence, and the revenue data to build predictions. So we provide a score 0 is that account will churn 100 is, the account will be retained.

1835
05:41:07.020 –> 05:41:09.750
Alok Shukla: But here comes the agent. Api comes into place.

1836
05:41:09.960 –> 05:41:37.630
Alok Shukla: Remember that for any kind of a revenue person, or a salesperson or customer success person. These numbers are just numbers unless explained by something else, so they generally what they do is when they see a bad score. They go across all sources of data to figure out why this is happening. What? What do? Let’s look at the revenue data. Let’s look at chat. Let’s look at gong. Let’s look at zoom calls. Let’s look at product usage data where the problem might be happening.

1837
05:41:38.130 –> 05:41:42.579
Alok Shukla: But in funny story agent does for you. If you look at this

1838
05:41:42.840 –> 05:41:48.170
Alok Shukla: color story has automatically calculated, created a summary and a diagnosis

1839
05:41:48.400 –> 05:41:51.890
Alok Shukla: for each and every account. Let me kind of take you to a specific account.

1840
05:41:52.280 –> 05:41:55.100
Alok Shukla: Oh, for Accenture!

1841
05:41:56.310 –> 05:42:02.429
Alok Shukla: So funny story on every given daily basis is analyzing all your accounts

1842
05:42:03.860 –> 05:42:11.339
Alok Shukla: and calculating across all your revenue data. Why, that account is behaving the way it is behaving.

1843
05:42:12.760 –> 05:42:20.579
Alok Shukla: So it’s so you basically say, probably 30% of your daily time that you spend on analyzing your accounts. You don’t have to do that

1844
05:42:20.690 –> 05:42:23.890
Alok Shukla: simple. The agent does it for you to look at this

1845
05:42:24.720 –> 05:42:30.569
Alok Shukla: data, analyzed from product sources, from support tickets from conversations

1846
05:42:30.690 –> 05:42:42.079
Alok Shukla: and also adding a summary. Why, that might be happening. It doesn’t stop there. It’s also picking up the data from business side. What? Maybe everything is good with your product. The business is not doing well.

1847
05:42:42.350 –> 05:42:55.540
Alok Shukla: It is also modeling on everyday basis. Why, that churn might be happening. What are the sources of risk. This is entirely agentic. It’s building the sense on its own. But now I come to the next thing which actually, which exceeds.

1848
05:42:56.220 –> 05:43:06.490
Alok Shukla: compared to any other tool in the industry which have not even thought about. This story calculates leading indicator of churn from your conversations, from your unstructured data.

1849
05:43:06.650 –> 05:43:09.479
Alok Shukla: So it is picking up what is called as needle movers.

1850
05:43:09.800 –> 05:43:13.169
Alok Shukla: Needle movers are kind of things. Think about it like

1851
05:43:13.460 –> 05:43:18.920
Alok Shukla: before a revenue event happens, whether it churn or expansion or upgrade.

1852
05:43:19.420 –> 05:43:24.869
Alok Shukla: People start to give you indication in their conversations, all of 3 to 6 more, 9 months in advance.

1853
05:43:24.980 –> 05:43:35.150
Alok Shukla: They will tell you by their feature, request the pricing discussion, the competition discussion, some hint that will drop if you catch them, you can get over the problem. If you don’t, you will be left behind.

1854
05:43:35.580 –> 05:43:57.320
Alok Shukla: And the problem with most of the account executives is that they don’t know where to look for that. And can they be on top of researching. All the data but 42 is doing is because it has modeled that its agent is automatically behind the scene is identifying those conversations which have been known to be preceding journal expansion events.

1855
05:43:57.480 –> 05:44:08.749
Alok Shukla: So, for example, in this case, it is saying, your, we saw that your customers were talking about competition. This is a risk, and they have been talking about repeated to. They have been dropping hints across multiple channels.

1856
05:44:09.060 –> 05:44:11.130
Alok Shukla: They were talking about some feature request.

1857
05:44:11.300 –> 05:44:13.539
Alok Shukla: They asked us, and their tone is changing.

1858
05:44:14.100 –> 05:44:27.810
Alok Shukla: They were talking. There were some people who were planning to leave the company, or there was some internal reassignment or some. There has been operation on some team where the team size might have been reduced. So they have been talking about it. Proof of the conversations.

1859
05:44:27.990 –> 05:44:33.070
Alok Shukla: or they’ve been concerned about the pricing of.

1860
05:44:34.180 –> 05:44:39.110
Alok Shukla: there might be a issue with the contract size. There might be unsustainable pricing model.

1861
05:44:39.320 –> 05:44:46.859
Alok Shukla: or there are proof points, or there are task that told you about something. There was a bug in the product.

1862
05:44:47.030 –> 05:45:03.410
Alok Shukla: so the the issue can be across anything. But the point here is, it’s in very hard for a 1 customer, success person, or one account executive to do all this research and get this kind of research served you on a table for each and every account of his

1863
05:45:03.540 –> 05:45:09.980
Alok Shukla: on everyday morning. Honestly does it for you, taking away a huge work lift, and that’s an agentic power of agents.

1864
05:45:10.740 –> 05:45:13.979
Alok Shukla: But then we go a step further. We don’t stop here.

1865
05:45:14.140 –> 05:45:29.200
Alok Shukla: So now the problem here is okay. You give me the data. But are you expecting me if I’m a customer, success person, to daily go into the product and read all of these things and figure out what other tasks I need to focus on. Should I talk about pricing? Should I talk about personal change, in which account

1866
05:45:29.500 –> 05:45:31.220
Alok Shukla: it can become overwhelming?

1867
05:45:31.500 –> 05:45:33.300
Alok Shukla: So funny story goes. The next step

1868
05:45:33.830 –> 05:45:54.739
Alok Shukla: automatically looks at your accounts and identifies what are the tasks that you should be doing? The 3rd level of Agent or 4th Level of Agent so far where it is now figuring out, okay, this is the important thing you need to do. This is an account. This is a churn list. The account is expiring. So this is important. And basically it tells you how you’re

1869
05:45:56.720 –> 05:46:00.179
Alok Shukla: so. Your entire day has been planned in advance

1870
05:46:00.520 –> 05:46:05.579
Alok Shukla: with is why that account is important, prioritize by priority low medium.

1871
05:46:05.780 –> 05:46:19.479
Alok Shukla: so that you, when you come in the morning, all your issues accounts have been analyzed. The needle movers have been calculated, then, further on, the task has been created. So the entire workflow has been created. The only thing that you do, only thing you focus on

1872
05:46:21.070 –> 05:46:25.430
Alok Shukla: exactly how you want to be right.

1873
05:46:26.070 –> 05:46:37.329
Alok Shukla: So once we have done that, that’s the thing we do, let’s come to the final. And one of the most powerful things on a story offers. So all of these things have out of box features

1874
05:46:37.440 –> 05:46:52.650
Alok Shukla: which means agents have been programmed to do certain things, build your data, build your systems. Analyze them. Create your support. Your do needle mover analysis create AI task and so forth and so on.

1875
05:46:53.200 –> 05:47:00.439
Alok Shukla: But now you can do open, ended analysis. And that’s where I was coming to. You should have seen a lot of tools that couldn’t do. Search over your data.

1876
05:47:00.550 –> 05:47:05.450
Alok Shukla: but we can. But now, if we follow through, you can do reasoning. Let me kind of give you a few examples.

1877
05:47:06.080 –> 05:47:11.060
Alok Shukla: One of the questions of a lot of each customer. Success people is that, hey? I want to know about feature request.

1878
05:47:11.650 –> 05:47:12.999
Alok Shukla: How do I go about it?

1879
05:47:14.630 –> 05:47:17.639
Alok Shukla: And you can ask a question. For example, give me.

1880
05:47:18.160 –> 05:47:31.180
Alok Shukla: So let’s say, a customer success person heard from a customer. They are looking for a feature request. You go to your product manager. Your product manager says, I’m not going to solve a 1 feature request for one customer. I want to build for market. Give me a broader use case.

1881
05:47:31.300 –> 05:47:42.600
Alok Shukla: How does one revenue person or one customer success person can even make that case. They don’t have capacity to go to all product data, all revenue data, all support tickets to even get that analysis.

1882
05:47:43.090 –> 05:47:44.950
Alok Shukla: That’s where funnel story comes in.

1883
05:47:45.410 –> 05:47:51.339
Alok Shukla: So let’s ask the question. Give me feature request across all accounts.

1884
05:47:52.910 –> 05:47:55.630
Alok Shukla: or last, let’s say, 30 years.

1885
05:48:09.550 –> 05:48:10.769
Alok Shukla: This is online.

1886
05:48:24.970 –> 05:48:34.350
Alok Shukla: So what suddenly happened was that you got all the feature requests. But the more interesting part is that you got the information about all the customers who asked for it.

1887
05:48:34.650 –> 05:48:41.710
Alok Shukla: and it is also telling you which are the con conversation as a proof point. So you go to your product manager says, I know

1888
05:48:42.120 –> 05:48:56.359
Alok Shukla: I now have. There are so many customers asking for the same feature request. Will you do now? There is a different use case. But let’s say, product manager, ask you, no, no, this data. I need it in terms of how many total number of people are asking for it.

1889
05:48:59.340 –> 05:49:00.500
Alok Shukla: Give me a second.

1890
05:49:08.360 –> 05:49:12.099
Alok Shukla: I want to know how many people have asked for these feature requests.

1891
05:49:12.700 –> 05:49:18.169
Alok Shukla: And I want to know what is the total revenue associated with each of these accounts.

1892
05:49:18.630 –> 05:49:21.390
Alok Shukla: So let’s say so. The question is

1893
05:49:22.410 –> 05:49:26.800
Alok Shukla: feature request showing different columns, number of customer asking for that feature request

1894
05:49:27.080 –> 05:49:38.340
Alok Shukla: customers, and a total aggregate revenue, because that will create a prior tension. How many people ask for the same thing? But how much more money is associated with that because that should drive the business. Let’s ask it.

1895
05:49:57.090 –> 05:50:08.979
Alok Shukla: Okay, we got this data. There is no other tool in the industry which can do this kind of analysis. Because now you are stitching data from your conversation with your product data, with your revenue data combining together in one thing.

1896
05:50:09.820 –> 05:50:20.760
Alok Shukla: But now you have to go to the next step. You’re saying I’m not going to. Just with data. I have to make a more detailed case, more strategic case. Why that matters. Okay, let’s ask a different question.

1897
05:50:20.970 –> 05:50:37.250
Alok Shukla: Create a memo. To describe the top themes of these feature requests that I can share with my colleagues, explain the example and references to companies and people asking for these requests try to estimate the Aggregate Revenue Company, asking for the feature request for each of these themes. So we are going even more deeper.

1898
05:50:38.070 –> 05:50:39.540
Alok Shukla: Let’s ask this question

1899
05:50:40.280 –> 05:50:52.269
Alok Shukla: at every level. You’re not just simply doing searching the data. You are doing a reasoning. You are developing a reasoned response that you can take, because once you have that level of memo along with the data along with the ask.

1900
05:50:52.640 –> 05:50:59.200
Alok Shukla: you have just completed a month work of a product management in 2 min.

1901
05:50:59.760 –> 05:51:00.950
Alok Shukla: So look at this.

1902
05:51:05.550 –> 05:51:06.740
Alok Shukla: that’s all for you.

1903
05:51:06.960 –> 05:51:15.279
Alok Shukla: So I mean, I can go deep, very deep into this thing. But I just wanted to kind of talk about that. How an agentic product.

1904
05:51:15.490 –> 05:51:30.609
Alok Shukla: which is designed for customer super intelligence looks like, and how many small small agents have been baked into the entire product to make it happen. And that’s why the theme. It’s 1 of the largest agent deployment in any product that has ever been seen.

1905
05:51:30.950 –> 05:51:34.089
Alok Shukla: With that I will stop here, and happy to take questions.

05:51:34.930 –> 05:51:44.609
Julia Nimchinski: That’s super impressive a log. And before we transition to the questions we received here from the audience, I’m just curious. Do you even have competitors in the space.

1907
05:51:45.150 –> 05:51:46.450
Alok Shukla: Sorry come again.

1908
05:51:46.860 –> 05:52:00.830
Julia Nimchinski: Yeah, before we transition to our questions with the audience. I’m just curious. This is really impressive and holistic use case for Cs, and honestly, I’m not sure that I ever seen like an end to end

1909
05:52:01.050 –> 05:52:08.590
Julia Nimchinski: Demo like this. So I’m just curious. Do you even have competitors specifically direct competitors in this space.

1910
05:52:08.590 –> 05:52:11.024
Alok Shukla: So I think in this space

1911
05:52:11.850 –> 05:52:27.860
Alok Shukla: we started with customer success. But we are beginning to compete with tools like lean copio which are into general AI. Search and space. We are also competing with new company like wisdom. AI. Many of the new vendors that are coming in.

1912
05:52:27.990 –> 05:52:41.049
Alok Shukla: I mean the super intelligence. Space is pretty new, and you’re right about it. So customer success is our entry point, where we always do. But I think our product can be used by anybody who cares about customer data.

1913
05:52:42.360 –> 05:52:44.559
Julia Nimchinski: How do you even define the category?

1914
05:52:45.500 –> 05:52:48.140
Alok Shukla: So I think I’m naming it. Customer super intelligence

1915
05:52:48.642 –> 05:52:55.907
Alok Shukla: the idea here is, think about the evolution of the business Intelligence platform which was used by multiple

1916
05:52:57.280 –> 05:53:02.819
Alok Shukla: products. But they were only doing the part of analyzing some of the data, especially structured data.

1917
05:53:03.170 –> 05:53:19.060
Alok Shukla: So there is a horizontal expansion. This sorry. There’s a vertical expansion in terms of that. Now we can analyze both structured and unstructured data with the power of conversational framework, but from a horizontal expansion. We not only analyze, but we can take actions.

1918
05:53:20.330 –> 05:53:30.769
Alok Shukla: So start, think about the bi which is growing in capabilities, but also expansion into action. Category data plus action together. So I will say

1919
05:53:30.910 –> 05:53:41.200
Alok Shukla: I, I’m not very particular about what category it comes into. It’s more about what kind of use case we are solving, and that’s what will matter to ultimately the customers

1920
05:53:41.400 –> 05:53:58.459
Alok Shukla: from a benefit perspective customers will see reduced churn customers will see actually, productivity. I know about couple of our customers who reported that in just 9 months they were able to handle twice or thrice the number of customers with the same number of people without burning out.

1921
05:53:59.310 –> 05:54:02.730
Alok Shukla: and that’s 1 million dollar dollars of saving. Please go ahead.

1922
05:54:03.490 –> 05:54:15.700
Julia Nimchinski: Definitely one of the questions we are seeing here, is, how do you ensure alignment between 100 plus AI agents and human Csms without conflict and confusion?

1923
05:54:16.600 –> 05:54:20.250
Alok Shukla: So that’s so one of that’s a great question. So

1924
05:54:20.790 –> 05:54:29.840
Alok Shukla: that’s where our background on how we design these things comes into the bank being so, all 3 co-founders actually comes from a cyber security background.

1925
05:54:29.990 –> 05:54:33.819
Julia Nimchinski: And one of the things that we have done in cyber security is that.

1926
05:54:33.950 –> 05:54:35.150
Alok Shukla: We

1927
05:54:35.990 –> 05:54:50.509
Alok Shukla: had to. And we’ve always had to work with not enough data, but still work with a high degree of precision. Because, remember, if cyber security people don’t do their job effectively. Airlines stop working. We have seen that in past

1928
05:54:50.760 –> 05:55:04.389
Alok Shukla: right? So that’s kind of training of design that we come to. So we have baked in that. How do we ensure that the quality of data is accurate? It’s secured and the outcome it produces are predictable.

1929
05:55:04.480 –> 05:55:28.959
Alok Shukla: So we so the agentic swarm actually has agents which produce the data. Then we have agents who test the data before actually showing it to anybody. But all the output you’re saying is actually going through multiple agents. So you are asking question to one agent which is throwing the data to some other agent which are doing some part of the analysis. The 3rd agent is taking the data for general printing it. 5th agent is quantifying it before it comes back to you

1930
05:55:29.190 –> 05:55:35.949
Alok Shukla: so short? That was a long answer. Short answer is, yes, it has been baked into the design in depth. That’s our design. Philosophy.

1931
05:55:37.540 –> 05:55:47.130
Julia Nimchinski: On this note. The next question is, what guardrails were essential to keep customer experience personal and not robotic.

1932
05:55:50.540 –> 05:56:05.589
Alok Shukla: That’s an interesting question. So I mean, personally, in the sense that agents they’re part of the product which is out of the box, which is more focused on enterprise workflow. So these are more standard for pretty much all people who use the system.

1933
05:56:05.780 –> 05:56:12.940
Alok Shukla: But, Felix, when you interact in a conversational format, where you where you want information or assign instruction that is personal.

1934
05:56:13.180 –> 05:56:16.680
Alok Shukla: that understands about what you are asking about. I mean, I have

1935
05:56:17.500 –> 05:56:22.309
Alok Shukla: customer success leaders or revenue leaders who are asking those questions that

1936
05:56:22.728 –> 05:56:34.359
Alok Shukla: that they talk to Felix in a very personal way. Remember, Sam Atman used to say, people talk a lot about. Thank you. And please, we see that a lot working with Felix, if that’s the question. If that is the intent of the question.

1937
05:56:36.040 –> 05:56:46.200
Julia Nimchinski: Look one of the questions here, and actually, a couple of them are focused on resistance internally in terms of adoption.

1938
05:56:46.360 –> 05:56:48.799
Julia Nimchinski: And how do you typically overcome it?

1939
05:56:49.580 –> 05:56:51.109
Alok Shukla: Sorry. Can you ask the question again.

1940
05:56:51.110 –> 05:57:00.719
Julia Nimchinski: Yeah, so what’s what are the biggest points of resistance in selling the software and adopting it

1941
05:57:01.510 –> 05:57:02.820
Julia Nimchinski: within your Icp.

1942
05:57:03.370 –> 05:57:09.340
Alok Shukla: So I think I, from a technology point of view, we are pretty comfortable. We are extremely war right now.

1943
05:57:11.210 –> 05:57:17.680
Alok Shukla: I think in general, and I am very understanding of this, that most companies.

1944
05:57:18.460 –> 05:57:24.260
Alok Shukla: most leaders, still do not know how to make the case for AI internally.

1945
05:57:25.184 –> 05:57:32.429
Alok Shukla: How do they justify investment. It’s 1 thing. See, most of the large companies that we talk about. AI are b 2 c.

1946
05:57:32.900 –> 05:57:40.110
Alok Shukla: where they are sold directly to individual, whether you’re looking for chat, gpt, or or cursor, or whatever the names might be.

1947
05:57:40.430 –> 05:57:52.829
Alok Shukla: But the moment you look at the enterprise the deployment is much harder, because, first, st not only you need to ensure that the AI is predictable, but what big changes you can actually guarantee.

1948
05:57:53.440 –> 05:57:54.460
Alok Shukla: And

1949
05:57:54.760 –> 05:58:04.869
Alok Shukla: my, a lot of my effort is actually helping the leaders make the financial case. So one of the things we launched couple of months back was an Roi and AI calculator

1950
05:58:04.990 –> 05:58:21.679
Alok Shukla: where we help them how to make the case, and what kind of things it can solve, and how they can successfully make the case. So that’s that’s my single biggest challenge is more on the sales side, less about on the product side, or an option side. If if that also makes sense.

1951
05:58:22.330 –> 05:58:27.029
Julia Nimchinski: Absolutely. And what’s next? In terms of your roadmap?

1952
05:58:27.530 –> 05:58:31.829
Julia Nimchinski: What are you most excited about? And what can you actually speak to.

1953
05:58:34.970 –> 05:58:38.902
Alok Shukla: So we actually are releasing features almost every month.

1954
05:58:39.590 –> 05:58:43.369
Alok Shukla: our core competency, as we said, is, is in the data architecture.

1955
05:58:43.590 –> 05:58:59.840
Alok Shukla: So we are driven by use cases. So one of the interesting use cases that is coming to us right now is, how can we? Now we have the data? Can we assign some of this task to some of the agents which can now start interacting with customers?

1956
05:59:00.070 –> 05:59:11.890
Alok Shukla: See, I know, I understand, that there’s a lot of excitement in the industry that agents can interact with the customers. But, as I said, my background is from cyber security, and I’m very conscious that

1957
05:59:12.140 –> 05:59:23.290
Alok Shukla: selling into enterprise. I know customers will be happy with less features, but they will want the any agent to be predictable if a single agent misbehaves

1958
05:59:23.630 –> 05:59:26.250
Alok Shukla: while interacting with the customers. That’s a

1959
05:59:26.410 –> 05:59:34.240
Alok Shukla: that’s a wrap on the company’s prestige, and also the wrap on the on the vendors prestige.

1960
05:59:34.360 –> 05:59:47.270
Alok Shukla: So we are being very careful. We. We are kind of rolling out small incremental change in that effect, mostly not because of technology, more, more to be to get the confidence of the buyers in that category.

1961
05:59:47.970 –> 05:59:59.749
Alok Shukla: So that’s basically is most of things on a roadmap mostly is to introduce agent that us autonomously interacting with customers in a manner that enterprise can adopt them.

1962
06:00:02.100 –> 06:00:06.689
Julia Nimchinski: Look as a thought leader in the space, and especially with your background.

1963
06:00:07.020 –> 06:00:13.330
Julia Nimchinski: Just curious your thoughts. On sort of intelligent transition

1964
06:00:13.580 –> 06:00:18.170
Julia Nimchinski: into a workflow just like your demo. Now.

1965
06:00:18.550 –> 06:00:25.930
Julia Nimchinski: because when we were preparing the summit, we asked all of our community members what topics there are actually problems there

1966
06:00:26.060 –> 06:00:32.729
Julia Nimchinski: focus on most now, and a lot of Cxos were specifically mentioning the transition piece.

1967
06:00:33.060 –> 06:00:39.210
Julia Nimchinski: So what would you advise to all the leaders listening to us now?

1968
06:00:39.390 –> 06:00:45.509
Julia Nimchinski: How do they transition from their existing tech stack to something like you showed us now.

1969
06:00:46.410 –> 06:00:52.670
Alok Shukla: See, I think, there are some areas where the value is straightforward.

1970
06:00:53.090 –> 06:00:58.500
Alok Shukla: For example, the data analysis, especially in some of see, let me kind of take a step back.

1971
06:01:00.370 –> 06:01:04.550
Alok Shukla: There are in enterprise data that is waiting to be harnessed

1972
06:01:05.473 –> 06:01:18.549
Alok Shukla: unstructured data has lived in the enterprise for a long time. There was no tool search already, is there? Reasoning is coming up? That’s an automatic case. You don’t have to make. That transition is easy.

1973
06:01:18.880 –> 06:01:25.219
Alok Shukla: So that is something I will say, go for it. Test the tools. You will get benefit of whatever you’re doing correctly

1974
06:01:26.020 –> 06:01:37.920
Alok Shukla: in terms of interacting with the customers. I would say that even though it kind of goes against my own interest, I would say, experience caution. Any interaction with the customer has to be predictable

1975
06:01:38.408 –> 06:01:53.469
Alok Shukla: try out in smaller use cases. Give cust. Give your customers enough warning that this is something that you are trying out with an agent, so that you allow both parties to settle down in that interaction without rolling out something big.

1976
06:01:53.540 –> 06:02:11.319
Alok Shukla: So there are some areas of faster transition. There are some areas for slower transition copilot are going to move faster. Direct agents are going to move slower. That’s would be my ex. That’s why, based on my experience interacting with almost 500 customer success leaders over the last one.

1977
06:02:12.400 –> 06:02:15.900
Julia Nimchinski: Yeah. And that’s actually my last question here.

1978
06:02:16.790 –> 06:02:24.329
Julia Nimchinski: There are a lot of community members with premium brands, more mid market enterprise brands, establishments.

1979
06:02:24.720 –> 06:02:28.119
Julia Nimchinski: and they’re really cautious about deploying something like this.

1980
06:02:28.330 –> 06:02:37.120
Julia Nimchinski: So the question is, how do you maintain your brand voice and generally reputation and experiment with AI.

1981
06:02:38.430 –> 06:02:56.919
Alok Shukla: Again, going back to that focus on work. First, st focus on co-pilot use cases more than the direct agent use cases. See? Ultimately, I will say that agents require good data to work. If your data is not good, the agent will act irresponsibly.

1982
06:02:57.340 –> 06:03:06.230
Alok Shukla: Agent cannot exist because they need some level of intelligence on which they will make decisions. If your underlying data is bad, you are going to have problems

1983
06:03:06.430 –> 06:03:25.229
Alok Shukla: irrespective, or whatever use case you do, that you will swarm people. You will touch people without they having any need for it. So I believe that is the area that that’s how you should prioritize. So focus on data, first, st focus on internal use cases. 1st focus on co-pilot. Use cases 1st that way, you will reduce the risk

1984
06:03:25.480 –> 06:03:29.230
Alok Shukla: and then go towards the bigger. Use cases as you go along.

1985
06:03:30.800 –> 06:03:32.039
Julia Nimchinski: Thank you so much. Your luck

1986
06:03:32.410 –> 06:03:40.360
Julia Nimchinski: really fascinating. Love the demo. And just for all of the folks watching. What is the next step.

1987
06:03:40.930 –> 06:03:44.640
Julia Nimchinski: What? How can we learn more about final story? Is it a demo?

1988
06:03:44.820 –> 06:03:48.470
Julia Nimchinski: Is there any pre premium experience to us more.

1989
06:03:49.190 –> 06:04:11.600
Alok Shukla: I think the best way to do is to come. So I’m I am always available to set up a demo for you, and I can explain to you while live. So go to my website, and there is a link for setting asking for a demo, or you can do it yourself, but I generally advise do it with me because I could help you out and connects with your enterprise user experience. And I promise you.

1990
06:04:11.770 –> 06:04:14.330
Alok Shukla: and I can actually put my word on this.

1991
06:04:15.080 –> 06:04:22.149
Alok Shukla: Be able to set your entire system, especially on the self. Serve Demo piece that we offer in 31 min

1992
06:04:24.290 –> 06:04:28.599
Alok Shukla: appointment of enterprise, revenue product in the industry period.

1993
06:04:28.600 –> 06:04:29.180
Julia Nimchinski: Oh!

1994
06:04:30.560 –> 06:04:32.979
Alok Shukla: It would be 9 months well spent. Thank you.

1995
06:04:33.830 –> 06:04:43.800
Julia Nimchinski: Super impressive. Thank you again, and that wraps up day 2 of the agenda AI summit join us tomorrow for day. 3.

1996
06:04:43.930 –> 06:04:50.960
Julia Nimchinski: The grand finale. There are a lot of exciting speakers, Cxos, Vcs analysts.

1997
06:04:51.200 –> 06:04:58.860
Julia Nimchinski: And yeah, just remember, we are building the stock market of skills. You can book one session

1998
06:04:59.030 –> 06:05:05.030
Julia Nimchinski: expert consultations coaching with majority of the speakers on this summit.

1999
06:05:05.390 –> 06:05:11.819
Julia Nimchinski: So yeah, see you tomorrow, and any final words, oh, I’ll leave it to you.

2000
06:05:13.242 –> 06:05:16.749
Alok Shukla: For me, I’m I’m pretty excited to be here. Thank you.

2001
06:05:17.320 –> 06:05:18.120
Julia Nimchinski: Thank you.

2002
06:05:18.240 –> 06:05:19.430
Alok Shukla: Thank you. See you soon.

2003
06:05:19.560 –> 06:05:20.090
Alok Shukla: Okay.