How Customer Journey Automation Drives 35% More Conversions with Dikshant Dave Zigment CEO

Clearly we do not think that you know AI agents have to replace humans and good thing is that most businesses that we talk to they are not looking at that either most business that we speak with they're saying okay now so this how can we do more of what we are doing so you know so at least that's a good sign that even the leaders are saying I have a team of you know [music] 40 people we are this is our throughput and like you know the first thing is that can we increase the throughput because I'm not interested in necessarily kind of cutting [music] the cost out. I want to kind of grow the business. >> Hi, my name is Demetri Bonichi and I'm a content creator, agency owner, and AI enthusiast. You're listening to the AI agents podcast [music] brought to you by

Jot Form and featuring our very own CEO and founder, Idkin Tank. This is the show where artificial intelligence meets innovation, productivity, and the tools shaping the future of work. Enjoy the show. Hello and welcome back to another episode of the AI agents podcast. In this episode we have the founder and CEO of Ziggman, Dshant Dave. How you doing today, Dan? >> Very well, Dimmitri. Thank you for having me here. How are you doing? >> Absolutely. I'm doing well. Uh very excited to chat with you about Ziggmet. Um very cool product. It seems like I I like the tagline on the top of your website. It says engage every customer and automate each workflow. That's exactly what every person with a business wants to do. So, um I feel like uh and as and as well at the top it does say that you guys are

um you know doing agentic customer journey automation which kind of spells that out a little bit more clear. But to hear it from you directly I'd love to hear two things. One um you know how did you get into the world of AI and then we'll go from there. >> Sure. Okay. Uh so u you know I uh with my previous uh startup uh you know which folded in 2022ish 2021 end uh and uh the co you know it was due to co and I took a sabbatical and I decided okay I'm going to kind of just you know pull my heels you know figure out get my bearings you know in by the time I started thinking that I want to do something uh it was you know the AI era had begun you know so Chad GBD was you know just about launching

and uh and and you know so the co-ound my co-founder karma and I we had been kind of trying out you know thinking about like the problems and stuff and uh there was a specific problem that we had in mind and you know it just coincided that we had uh openai at launch chat GPD and that kind of changed and gave it a lot of perspective that okay there is now uh elegant elegant way of you know building a solution that we were thinking of you know solving. >> Okay. Very cool. And you know you are somebody who started your company in 2023 is that correct? Fairly recently. >> That's right. Yes. >> Yeah. Okay. And you see a lot of so-called rappers in the space and whatnot, but you specifically built out this AI um agentic customer journey platform. What was the specific like

gap that you feel like you saw in the market where standard chat bots were kind of failing and what was that light bulb moment where you realized the tech had finally matured enough to move from chat to um where you're at? >> Yeah, I think you know a great question and this has been you know I would say that like this has not been a like a light bulb moment you know it has more >> throughout. Yeah, exactly. And uh so while we started at uh you know we wanted to figure out like what can we you know uh so I'll kind of just kind of tell you about the problem that we had in you know mind uh in my previous startup we were doing uh supplements you know health supplements and for the US market and we' built uh a pretty elegant algorithm

and you know it was all online we were not converting enough we were not converting to the extent we would like to do and although people were kind of taking this 28 question, you know, test, you know, we we kind of had a completion rates of 71, you know, percentage on this test and yet uh, you know, very poor conversions and, you know, that made us kind of again the founder hustle. We started kind of calling up these people, emailing them and as it would happen that eight out of 10 of these customers would kind of you know convert you know once we had a kind of know you know engaged you know with them asked them what what was it that was blocking them and uh the thing was that everyone had a very different reason to kind of not go with you know

with with the purchase. uh although they were pretty interested in the solution or the problem was kind of connecting with them but there there was some there were few things that were not uh you know they were not getting answers for and we tried kind of doing a lot of these you know these things solving through you know the online FAQs and stuff but it it just did not it it probably we kind of figured figured that it required a consultation you know and and consultation at a highest level. We also tried you know solving it by putting a small call center but uh that did not solve uh and we figured that okay maybe our level of conviction a founder level of conviction or a very high level expertise was required to you know uh solve that problem. uh fast forward you know 3

four years ahead uh this problem had stayed that oh I wish I could replicate myself you know uh and you know have you know engage these uh you know customers or prospects and possibly convert so we were starting to kind of we started as you know Ziggman started as okay let's solve this problem can we build uh sales agents uh which had like this whole angle of persuasion so not just support and and you know primary difference and as I educated myself that in support it's the customer who leads the conversation if it's sales then the saleserson kind of you know leads the conversation and that's like a pretty fundamental difference in how conversation flow in support versus sales and we were focused that okay we are building a sales agent so while we started as conversational sales AI u figured that the problems did

not end there uh this whole solution had to fit into a much larger tech stack you know within the marketing or kind of a growth organization and we would often kind of hear that okay this is working fine but I mean you know it's it needs to kind of you know be fully integrated with our CRM and and it kind of needs to kind of send this information to our CDP you know or kind of you know different pieces of text stack that they would have a drip email and stuff and as it happened we just kept solving these problems We kind of figured that it's not a silo, you know, software. It needs to be more of a platform or like, you know, something that integrates into the, you know, a system that a business has. Uh, and that kind of, you know, kind

of that was a journey from a conversational chat agent to, you know, a a journey platform. So, sorry about the long- winded answer for your >> Oh, no, that's totally okay. No, that's okay. You're fine. Um, don't worry about that. I that was a that was some good context and I I definitely think it was um it's fair fair to say like things don't always come as light bulbs. So um that was a pretty good um talk you had there for our listeners who are a little bit like tired of um buzzwords, right? The whole shows AI agents, agentic layers, I don't know what the heck that means. I mean I do but I'm being facitious. Um, can you explain in plain English like the difference between a standard conversational bot and the agents that you were building if you have like obviously you're going

to have these conversations with people that you sell. So something to that effect. >> Yeah. Uh yeah, I mean very fundamental question. You know in simplest word I would say you know the difference is autonomy. So our conventional chat bots versus the new AI agents that we talk about is autonomy. Now what what does autonomy mean? I think is uh where uh that piece of software which we call you know the AI or the agent can make decisions that you have not necessarily kind of scripted or tested for. So you know basis like you know if you have if we have a sales you know person who we hired and we gave you know that person all the material you know about the business and the products and stuff and we kind of said okay this is what this our general script is now go

and sell and every customer is going to be different but you're not going to train them for very specific type of customer you're just going to rely on here is our information you know here are some basic you know guardrails and just now go and uh that's the autonomy right so you've given autonomy to to that you know agent or a sales executive now in the previous version of our chat bots uh this was not the case there was no autonomy they had to be scripted and uh you know in they they were like decision trees that okay give them like ask them this question and then if they say this then kind of this is your tree you know you know branch that you're going to follow if they say know that kind of follow this path and then the whole decision kind of

in a tree continues where you are deciding what that conversation you know uh how the conversation would flow the problem with that I mean and we all are aware of that that the experience you know terribly you know like you know it was just terrible right and you know there was uh people would kind of not you know uh would not want to engage with these chat bots and couldn't wait for a human to kind of be at the other Uh it was primary because it was such a scripted thing. It was you did not like customer could not kind of decide how it how he or she wanted to kind of you know get the questions answered. It was the chat script. Now in in the current form that where we call you know AI agents uh all that you're doing is training them

with with the information and you know putting the guardrails you know deciding on the basic script. this is this is the philosophy, this is how we approach our business and you're letting the AI agent kind of do the job. So I would say like yeah that's that's a primary difference. >> Okay. Yeah. I mean with with doing the job I think it's a fair question just as a followup there like where do you think these agents kind of really can sit in the stack of quote a team, right? because um maybe some people are going to think that it's not necessarily possible for them to do real team work. Um it is kind of contextual based on the type of roles we're talking about. Where do you think they can really sit on a team and can they even replace people in some contexts? Yeah,

I mean you know we have seen that I and I I guess you know first up like I don't think that AI agents you know today there is a deployment which kind of does you know entire task on its own like know without in a silo there is a human you know there is going to be a human in the chain somewhere either before that that particular starting point or after the end point where the agent kind of stops um In practically every deployments that we have done at Ziggment uh you know there is there there are humans who are kind of you know engaged before and then there are humans engaged after also. U to your specification what kind of tasks that we accomplish. Usually here we say that there are things you know in typical sales and marketing kind of roles. uh there

are so many uh repetitive tasks that kind of exist today and and you know uh you know just for example that okay if if there is uh you are an online store you know somebody kind of walks in into your Instagram page and leaves a message that is somebody who's going to be responsible for responding to that particular comment and and then kind of you know asking them like questions and trying to hoping that okay you you may drive the conversation towards uh having them purchase you know make a purchase Now u those are the tasks and then you know when you look at like you know teams that do it uh they're doing it like you know day in day out and you know uh the weekends it stops but again it starts on the weeks. So typically these tasks which happen to be

in between these things which are like okay like engage with customers either send them nudges reminders uh somebody if if you are in lead business and if you have you know meeting setup somebody's job is to kind of you know make sure that reminders are going up and confirming kind of you know the participation etc. Uh these are very repetitive tasks you know nobody enjoys or very few people enjoy doing these tasks. So these are the the prime things that that AI agent you know uh AI agents are good at. I would say that uh >> the closure of business like you know when if you are let's say if you're into build selling a high ticket item uh >> at the end of it the customer is going to want to kind of have a real human kind of shake hands you know uh

and then close the deal. So that's where the human would be required to make negotiation kind of you know uh explain the finer parts but everything before that you know at least a few steps before that is where AI can shine. >> Yeah that makes sense. Um yeah it's fair. I think there's a lot of different misconceptions right now in the world of AI. I I've dealt with a lot of them on this show like trying to address it with people individually outside of the show after learning from you know people like yourself. Um, and I do think it that's that's a fair assessment of kind of um where you're at. How does your system handle the um sentiment like aware um actions um that you talk about on your website? I I I saw it on there. I thought it was interesting. I think

sentiment is a big thing with AI right now. Um and trying to make sure that it's constantly factoring that in. How does it handle that? Yeah, I I think u I guess this is again one of the key factors that differentiates you know the AI agents of this current you know era generation versus the previous chat bots. Previous chatbots were basically uh stripped of all these software signals which was sentiment mood kind of you know detecting anger or frustration or anything. there was just no way because you were making customers just choose like given from a given three options you know so >> sure >> in in this current like you know with with the current gen AI agents because there is a human level kind of natural language conversation you often have like a much wider or kind of a broad uh conversations happening

uh there could be stuff that people talk about that you would not have imagined you know that would not have you know thought about when you're designing the AI agent but uh with the right information and right guardrails AI agents can kind of now converse and what happens there is that because you have all this data and information from conversation like platform like ours we detect you know there is a constant monitoring of all conversations and these signals keep getting updated on the dashboard in the back end that okay like you know there is uh you know strong intent detected uh urgency see kind of you know the user is looking at like uh you know urgent or like you know short-term delivery. So somebody asking that hey like this is I know this is like last moment but can I get uh package before

the labor day you know for example now he's not asking for that I need it now or urgency but from the conversation like like us humans can detect that oh this if we manage to kind of send it before Labor Day then we have a sale here. uh so these signals kind of get updated in the in the dashboard behind uh we are able to do this because all the conversations are now kind of tracked and analyzed and AI is kind of know uh is is is monitoring the you know uh the conversations and kind of know pulling out all these signals from the conversation. So, so, so that's and so you know the sentiment, mood, you know, urgency are all part of, you know, uh, these signals and >> and where do you think like the capabilities of AI was able to start picking

up on these things a little bit better? You've kind of been in the business. Um, you know, 3 years is a good amount of time because I think that was pretty much right at that would have been the spring after GPT3. >> Yeah. >> And 3.5 kind of like stormed the universe and it took the world by storm. So you kind of were at the beginning of capabilities. When did it kind of if you like start to change that this sentiment analysis was possible by um models? >> So to be honest I mean we started figuring this out uh even in 3.5 and uh 3.5 had a lot of error rate. uh what that made us was like you know build our side of guardrails and the platform much stronger that okay how do we rely less on the you know the AI and kind

of build our guard rails uh so that just kind of helped us you know build way more sophisticated systems at our end uh and you know like deploying in the businesses that has helped because we are now way more confident you know without relying on AI's goodness uh so to speak so u I would say that 3.5 even we build it I think we saw real difference in the qualitative you know uh output from from LLM was uh open AAI 4 so when four was launched I think we we saw that there was market difference you know in how it processed I don't think that anything has been that dramatic post that you know post that event so 3.5 to 4 was a big jump a very big jump but uh and things have been since then, >> you know. What about like the reasoning

models that like came out like I guess earlier this year? Did that have any sort of impact at all? Or even the I guess small reasoning models like GPT 4.1's technically a reasoning model. It's just small. >> Yeah. So I mean we've uh we've not you know we we did not need reasoning models most of the time. So conversational kind of you know uh because reasoning models take time uh in conversations you know in time responses are important right. So we kind of non-reasoning models but you know for conversational part and for most you know pieces where you know you don't require like a elaborate reasoning or analysis u you know normal 4 have been kind of pretty decent. Okay, that's fair. No, I I think um it's interesting because obviously you're going for just good responses quickly and it is fair to say in

my opinion like there is so much value in what the reasoning models do, but maybe that value doesn't come at a quick enough manner and maybe the value is not as outstanding as needed for it to warrant a upgrade. Um, like for example, I feel like even when like five came out, like GBT 5, a lot of people were like annoyed with the responses versus what four was doing, 40 so to speak. So it's like just because it's later doesn't mean in all context it's better. I think it's a decently fair assessment, wouldn't you say? >> Yeah. Yeah, I I agree. And I think uh there have been times when we used the reasoning model like you know one deployment where we required or in during conversation uh that the customers would send a document and that had to be analyzed you know bit and

for that we used uh the reasoning model you know for for analysis of that document that got you know the user who sent and we had a liberty to kind of you know uh tell that user that okay give me a moment I'll kind of let me analyze right so it's it's clearly a break in conversation and we could do that. So uh but yeah I mean there have been instances but uh for normal conversations yeah I mean standard stuff is good. >> Yeah. Okay. Um when you kind of are talking to potential customers right when you pitch this to businesses are they hesitant at all to hand over like closing capabilities let's say to AI? Oh yeah. I think uh you know u and I think we do not advocate that either. So we kind of think that you're you know let's just address

the uh you know the acutest pain that you have today and and which could be that most businesses to be honest is like oh we're getting so many leads you know identifying the five out of those hundred which are just ripe for my best people to focus on. uh that's a universal problem. So we kind of go in with that. Okay, let us help you kind of you know distill your 100 kind of bring that to like you know five and if not five maybe you know 10 15 and then reduce that set where you know that okay your best people are deployed on on the kind of final five and uh we will prep these leads kind of you know make sure that you know everything is done so that your best people would be you know so uh just focus on these uh

ready to buy or kind of you know very very warm uh conversations. >> Yeah. No, I think that's a fair distinction. Um there's there's sometimes when um people But have actually I am curious though. I know most of the time they're hesitant. Has anybody kind of come to you though with an expectation that it could do that? >> And then Yeah. Then they want the opposite, right? >> On very early on where AI was this, you know, took took the world by storm and you know, literally >> so an esoteric kuna heck knows what it is sort of. Yeah. >> Yeah. I was like okay let's just sell sell sell us you know sell these courses like you know I don't want to have any any human kind of doing anything and we managed to do that actually. So we we did kind of you

know did a complete sale uh on on chat. Um so yeah I mean I mean there have been cases but I think as we've kind of gone into the larger organizations it has been like okay there is this workflow and there are these pieces of the workflow. So there are like let's say there are these 20 things or 20 20 workflows that happen and we start with one or two and then kind of know move that okay like now AI is handling you know 10 out of those 20 workflows that that organization kind of did and uh you know it's it's just basically building confidence you know uh internally and externally or not expanding that. >> Yeah that's that's totally fair. Um, so you based on like kind of what I saw you you hit around um I think based this I I think this

is public knowledge is one of those things where you try to like uh get in phone businesses. You've kind of reached a really solid uh number of a couple million in revenue recently. um for for such a like young company um what was like the hardest hurdle that you had to solve to get the AI good enough let's so to speak to generate that kind of revenue and and buzz about your product >> um yeah I think I mean right from the beginning we have been very focused on delivering value and even if it kind of came at a cost of scale and growth uh we've been focused on that if the customer that we on board and we are confident of doing something if ultimately if you're not able to kind of deliver uh there's no business here and there's no company you know

that that should exist uh so we kind of even with the early customers we just took time we just made made sure that okay I mean it this has to work this has to work for you not for us like not our metrics they it you know uh AI has to impact your metrics you know so that focus build that DNA in the company where everybody talked about like hey you know what like you know this is happening internally we see that like our team saying oh we are hitting AI is hitting like good metrics but the eventual conversions are not happening and just the DNA that concern that okay like there is something a miss and we have to escalate this that maybe there is some issue with their call center or their kind of you know the different pieces that we are not

kind of handling just has given I would say like a better culture to kind of you just address uh customer problems way better and I think u once we are able to deliver that value each practically each customer is kind of you know like you it's a GDM in itself like some of the large customers we know that they operate in 100 countries you know so so we know that I mean you know we if we do it well we we do not have to kind of go out anywhere you know looking for business >> yeah okay that's that's very cool and you operate across web chat like SMS email and social media. So, like which channel um do you feel like is uh the hardest to to master right now? >> I think they're practically same. Each one has their own uh nuances and

I think this was like a maybe the first I would say few quarters kind of a concern that oh like oh email has threads and like you know like how do you kind of maintain threads because uh when you have a messaging conversation it's the same thread like we would be on on same thread all the time talking all the time but in email that could be like you know multiple emails just between two of us right so how do you kind of manage those things so but those are all uh problems were there that that were there and and solved now. So those are not the things that kind of know keep us up in the night anymore. >> What does keep you up at night? Now >> interesting question. I think u a few few things I guess uh one of the things

is of course u you know every business comes with its own set of problems which are very different. Uh they're all kind of connected but uh they're all different in their own ways. So we internally keep have been kind of building like okay what's the common commonality we can't be kind of reinventing the wheel what's the commonality between these problems so uh so yeah so things like these you know so things like you know like uh two years from now what kind of organization are we going to be and like you know how efficient you know uh we are going to be at solving these >> interesting uh what would you say is the company's like main ethos and like goals over the next few years. Uh so I think u ethos is is basically like you know just very customercentric very solutions you know

focused internally when we started uh very early on we decided that we do not want to be a product company like product let it be the outcome of what we do u and and having run few you know startups earlier I know that you know for a founder it's very easy to kind of start getting then romancing your product and then seeing that okay this is the solution for everything in in in the world. >> Yeah. That's how people are when they Yeah. >> And yeah, I mean it you can't help it. I mean it happens, right? Because you kind of know literally nurturing that. And I mean I was at least very clear that we ultimately that has to be a you know solving problems. And so we said okay we are a solutions company and product I mean product is a mindset. We

you know karma and I both are you know have product backgrounds. He's uh he's is tech. I can kind of come from the product side. So we say product will emerge like no matter what we will not you know let it you know slide but focus of the company should be solving problems and if some solution requires just a spreadsheet like in a Google sheet that solves a problem let's just suggest that let's not kind of sell them something you know fancy because we have so so yeah so that's that's the ethos and I think goal for the next few years I guess is just continue to do what we're doing I think one of the things that we Now because we've spent so much time and kind of you know mastering our conversations like getting them better and better and we kind of felt

earlier when there were combination that oh there are so many agent like know chatbot players today like there are so many conversational agents more than you can count >> uh >> we were clear that ultimately at some point in time somebody's going to say there is a difference in like you know there is a good agent and there is not so like good agent like you know conversation wise just like we have like not all sales people are same you know there are some people who you would like you know make you eat out of their hand and then there are some people who you you would avoid or or kind of be indifferent to. So we kind of knew that there has to be this conversational part is going to come in and that is now starting to show. So one of the things

that we kind of go to a lot of customers when we demo you know a lot of them say that yeah I mean this we've not seen this quality you know whether it's voice whether it's chat so so yeah so I think just focusing on that you know and and continue to kind of do you know just simple things you know no nothing too big. >> Gotcha. Yeah. No um that makes a lot of sense. Um, so another question I kind of have is just based off some of your claims on the website. So like your platform talks about how it can boost conversions by let's say 35%. Is it like purely speed to lead or is it the AI actually like better at qualifying than a human SDR for example? >> So thanks for asking this and then we have a a published case

study with Meta now. So meta has published this case study with our case study with no we have and where you know these numbers have been validated where where customer kind of know has accepted and we've kind of you know proven that this is happening u Nova is is Noah healthcare is one of our we started with them as a pilot and it's now more than a year now that we are running with them you know it's it's a full-blown commercial project uh with them and we've delivered this. So to answer your question, these conversions are actually end conversions. So they have a parallel system where Ziggment is not a part of it. They're running the same kind of you know the ad campaigns and then they have their human team kind of doing what they were doing and then there's Ziggment and you know

month there has been like you know market difference that we are at about you know between 35 to 40%. uh you know better conversions and and better conversions not just at at kind of you know conversational level but even at kind of the end conversion like really like money conversion uh so so yeah so we we've kind of you know been able to do that and I think that comes from something that I said earlier that uh just our focus on like let's let's make sure that like you know this delivers value and we know it delivers we got to kind of just be at it. >> Yeah. No, that that that makes a lot of sense. Um, kind of just furthering this a little bit more, what are some of the other kind of like case studies that you've uh done or any other

examples that you could give us so we can get some more color on how you've helped specific uh companies? >> Yeah. So, uh, Bajage Auto is is another customer we are doing. you know we started with them on Spain uh you know uh Spain market and uh their similar numbers are showing up uh it's been like what two months or maybe a little over two months now and uh our our metrics are you know right now at least 23% you know better than what they were doing earlier you know or through other channels so uh so that's happening uh we have we have many customers who We are uh live with company called script box where we kind of you know have delivered uh you know again the whole better journey experience you know that convert you know uh nurturing converting you know closing even

closing in some cases u you know we market difference where they've accepted yeah this is this is definitely much much better to a tune of like 25 30%. Uh there was another company where we had again a case study kind of you know printed out. uh this company was uh an online stock broking firm and uh yeah and then we were kind of converting so and they had a problem of uh their onboarding was like a so financial onboarding is a little complicated and especially in India where there are many uh you know checks and balances that happen you know there's KYC and then there's a you have to kind of attach your government ID and that kind of goes for verification uh it's a pretty long drawn and like multiple steps almost like nine steps that you take. Uh they had a problem that

you know the customers uh even after kind of you know agreeing to kind of uh downloading their app uh only 10% of customers were actually completing the entire you know verification test and uh and then with us uh that number was 24%. So, so, so, so, yeah, so like you know, we more than doubled their conversion. So, uh, so yeah, many cases like these and and as I said, I again like we we've relied on our, you know, ability that it has to make a business impact. If we don't, then you know, probably either we are not fit or maybe the business is not ready and like we should kind of take that cognition into cognizance. >> No, that's that's like that's a fair um I I think you should that makes sense. um kind of kind of moving forward with this a little bit.

I just don't want to talk about the market in general and we kind of talk generally about like what you guys can do from an agent standpoint and how much it can replace stuff, right? Um where do you think AI agents are heading in the next 12 24 months from a capability standpoint? My honest opinion is that we've kind of reached a plateau of the inherent ability of like you know the LLMs. I think that we are going to see like a much much gradual increase from here on. But having said that we it they are far more capable and now what we are seeing is that the actual builders and and developers and and smart folks are starting to kind of you know figure it out and utilize them in in ways that we it it you know that have not been done so

far. So in the last three years it's been kind of you know the whole euphoria around the LLM's capability and then some lowhanging fruit that okay like yeah I mean the whole rapper age right that 2023 and 2024 that you could build a rapper and kind of make millions out of it uh just because the LLM itself was you know very powerful but uh now that we are in in this where people are figuring out like how to kind of you know really utilize and like you know use it in ways that has not been So, so I would say that yeah, I mean 24 months we might see a lot of you know little bit things here and there but I think as is today like five and and kind of in the newest claw and everything they are like they're great. I think

you know they're far more smart and stuff. >> What do you think about the um kind of the MCP trend that's going on? the capabilities I should say >> internally we are bullish that I guess you know we are going to be in that age of like where there's going to be interoperability is going to be you know quite fundamental quite crucial uh we will see u the adoption has been on and off I think u I would still say that people like we are still in the age where businesses need a proof that this is going to make an impact and Uh that's why it's like before even kind of doing any complex stuff we you know the whole point is to kind of show at a business level that this on its own or like some simple use cases are working like real

value somewhere and then you can think of like building out very complex uh stuff. So I would say like MCP is great and then you know maybe even parallel kind of architectures will come up. There are some, you know, things with MCP. They'll only get better with, you know, more people coming in. But I guess the real challenge is that tech is a small part of that problem. The bigger part is that our business is ready to kind of have this complex mesh of you know interoperating processes and real you know you know units in their business and I think that's a harder uh part. It is going to take time is what my uh reading is. It is going to happen. As I say that you know we overestimate you know the impact in in the near term and grossly underestimate the impact in

the long term. >> The long term. >> This probably is a tight case. >> Do do you think that like the um capabilities may come at a decently quick rate but maybe the um adoption is going to be something that's going to struggle to follow? um >> like effectively effective effective adoption. I don't mean like adoption generally because there's a lot of people who use AI. >> That's rude but you know what I mean by >> no [snorts] and I think we all know that I mean I guess uh you know everybody is using AI now. So so that spread is going to happen. uh but yeah I mean the real use I think is going to take time and because there is uh you know the businesses are a very complex you know ecosystems of of things right and they are there is a

certain balance that is achieved in those ecosystems u you know moment something like this happens it's it's kind of a you know like almost like you're going to create imbalance and like so the immediate response of lot of people in the system is like is it worth it and so uh this needs needs to kind of show that 10x you know metric uplift in some something for people to say okay like maybe if this is the case that it's the battle is worth it because u a lot of people we see the champions you know in the organization our champions who are kind of you know like uh really like great help for us within kind of spreading the word within the organization and making that adoption happen. I think the pressure on them is so high like much more than you know what we

also kind of you know uh imagine because here they are putting their kind of job online and fighting for something that you know that is because nobody has seen it like they are just taking a a chance or or kind of faith on something. So, so I guess yeah, the adoption is going to be, you know, the the the uh it's going to take time, but we will see. Again, as I said, like in the long term, it may just be so much bigger than what we can actually, you know, imagine right now. >> Yeah, I do think that's a fair like assessment, right? Like sometimes what we'll try to do in these situations, we'll try to like make a prediction. Um, and I think a lot of this, practically speaking, on my show, um, or on this show has been conjecture. Right now, I'm

not saying it's going to be all wrong, but it has been conjecture. Um, but I do like hearing from people who are like actively working in it. Um, because I I'd like to I think that people want to have answers to questions like, and I'm curious your thought in five years, do you think an entire entry level like SDR role could be replaced by AI in theory? Um I think the SDR is uh I would say the hardest one because >> okay many nuances. U so I would say that like in the spectrum of uh if you just say the marketing spectrum the customer interaction spectrum I think the starting point and the end points in my opinion again uh and I do not have much basis other than my own experience. Yeah, sure. >> Starting point and the end point are the ones which

are best served by humans. Uh probably in between once the relationship is set like you can kind of you know automate a lot of you know there's an easy space for AI to get in uh and and then kind of a closure requires again the human element. So my personal thesis is that like you know SDR is why we've seen a lot of SDR kind of you know hype. I say hi because I am yet to see any company who's kind of you know told me that yeah I mean this like we've literally replaced everything you know uh with the opening SDRs or like agents which are like just generating new businesses like that's I have not I'm yet to see that. >> Yeah. Yeah. It's fair. I I think I would agree with you. I don't think it's going to be in a position

to where it's like capable of doing it. Um, I do think there's a level of humanity that will be [snorts] needed in in work in general. Um, like you're not one of these people that thinks, not saying that you it's an incorrect opinion because obviously we don't know. It seems like you're clearly in the camp of like we're going to keep working. There's not really going to be a mass replacement in the market and like no one's going to be working by AI obviously. >> Um, no. Yeah. I mean, I I do not think so. uh I think uh humans are far smarter than what we you know take them for and I think uh just collectively uh and individually not speaking individually but just as a collectively as a community I think uh we have survived and I think uh immediately there are things

that will kind of crop up where uh there could be a near-term kind of jolts uh but in the long run we will have like there will be coexistence it's just like a tool that's coming in much severe or a larger impact than earlier you know uh tech evolutionary kind of points but uh I think we will figure it out as a business or our kind of community. >> Sure. >> No, that's fair. >> Yeah. And clearly we do not think that you know AI agents have to replace humans and good thing is that most businesses that we talk to they are not looking at that either. uh most businesses that we speak with they're saying okay now so this uh how can we do more of what we are doing so you know so at least that's a good sign that even the leaders

are saying I have a team of you know 40 people uh we are this is our throughput and like you know the first thing is that can we increase the throughput because I'm not interested in necessarily kind of cutting the cost out I want to kind of grow the business so so yeah >> yeah a lot of I mean if people are focused on growing the business I do think it's fair that like that's that's a much cooler and better approach to kind of um to take it from cuz it it obviously gets concerning when people are just like cost cut cost cut cost cut cost cut but I do think a lot of people are more focused on how can I make my company better serve my customers kind of like how what you were showcasing and a lot of times AI just practically

is doing a really good job at doing that effectively and that doesn't mean you have to cut out a bunch of people at your your company it just means that you um are able to get more business maybe and have your current staff allocate less time to monotonous tasks and whatnot. >> Yeah, and that's a great point. I think you know that's a great way to look at it that use AI to kind of know improve the customer experience. I mean like you know there is so much you can do and uh you know uh that goes a long way in kind of you know just growing your business because a happy customer is going to be a valuable customer eventually. >> Absolutely. What do you think is your favorite um I guess I would say personal tool that you use that's not really obviously

you know your business is going to be your favorite tool we know we know this um but what is your favorite personal tool for AI use >> uh AI use only >> I mean sorry I don't mean AI sorry your person your personal favorite AI tool just for day-to-day use it could be personal use it could be like business use I'll give you my examples I really like text cortex because It's a um like an it's an LLM tool that allows you to make like writing agents that are really good that work in marketing and I really like a tool called Crisp AI which allows you to um have the background noise get removed and there is a slight audio delay but it's an AI like analyze what's the problem and then fix it sort of thing. >> Okay. Uh this keeps changing uh you

know so you know whatever it is there >> next week it'll be different we know but that's okay. >> Yeah. Yeah. U so I think lovable was one uh you know a few months ago and then I think I'm starting to kind of love Gemini much more. So Gemini kind of Okay. >> I'm I'm finding and I was kind of advocating inside also that like you know this this responses and like everything that I'm able to do with Gemini is so much better. Uh and this is non non-coding stuff. uh you know so uh but but more analysis like whether it's image creation, editing uh just generally creating documents and stuff uh so yeah so Gemini Gemini is my you know current love uh you know if you ask me >> and did you say lovable as well >> lovable was and then kind of

I kind of feel that yeah I mean it's it's not kind of improved from a point I saw or kind of I kept using it and like >> yeah I I think It's it's it's kind of limited in some respects from a backend perspective uh due to its disconnect. But they did release um lovable cloud which makes the edge functions and whatnot work natively so it's easier to debug stuff. Um and then that also goes on top of it with uh you know as the models get better I'm sure it'll get better. Like 4.5 Opus dropped yesterday and I played I've been playing around with lovable and it's really good. Like the >> Oh, really? >> Yeah. like there's a lot less circular problems like when you're trying to fix it and try to fix it and try to fix it. It I think between

it having the edge functions and storage built into the platform rather than it shipping it to superbase. I do think it's practically like doing a way better job. Um plus that 4.5 upgrade was really good. But just for your own personal if you want to check it out. Um >> yeah. Yeah. Yeah. I will. I think one of the notebook mentions is also notebook LLM. I'm I'm fine. >> I've heard it's good. I haven't really messed around with it. You like it? >> I love it. I mean, just making sense of complex documents and stuff. I you know, that's a tab that's always open in my machine. >> Absolutely. So, with that being said, I did want to just ask one last question. Where can everybody go to find you? Tell me a little bit about um or tell them where your website is and

make sure that you plug as many things as possible before we end the show. Sure, thank you for asking me. U so u our website is uh you know ziggment ziggme- n.ai and uh you know I am on LinkedIn my name is digshant digshant dave you know you can find me or you can email me at digshantzigment.ai that is d i k s h a n t at ziggment.ai AI. Um, I'm I'm pretty kind of I'm kind of saying a lot of things on LinkedIn, so that's the best place for uh to follow me or kind of connect. >> Well, I appreciate you for letting us know and make sure everyone to uh check out him over on LinkedIn and also check out everything that they are doing at Ziggmmit AI. That is spelled Zig N.AI. And if you like this episode, please everyone make

sure to leave a like, comment, subscribe, do all those good things. And also, please make sure to leave a review on Apple Podcast and Spotify podcast, including you, Dick Sean. It'll help out a lot with uh the algorithm and whatnot. So, everyone, please, please, thank you all the words, and we'll see you in the next one. Bye.