always, And so, versus, Market Office Hour, we will see how artificial intelligence will help marketing and sales to decide which is the best action to do next. I will be with Paul H.S. from Cypher, whose solution will help us calculate a scoring based on AI, a scoring that will be always accurate, always based on the best factors that engage and convert your opportunities. And Cypher will propose the next best action, whether it's for the marketing or it's for the sales. We'll see that together during this Market Office Hour. Hi everyone, we are today with Paul from Cypher. We met a while ago about Market Office Hour on the scoring with AI and today we will see the next step of Cypher, which is the next best action. Paul, I will let you introduce yourself and maybe show us a demo of your super app. Sure. Thank you, Sylvain. Nice to be here again. I'm Paul. I'm one of the co-founders of Cypher. Basically, we are providing an AI lead scoring solution that will plug to your sales and marketing tools. And the goal is easy. We just want to analyze all of your pipeline to identify high potential leads to convert so you can optimize your SDR efforts or your marketing efforts too. Maybe I could quickly share my screen to... Yeah, sure. Okay. So in this... Here is my Salesforce demo account. I'm going to pretend that I am a sales manager. That was what I was before, before building Cypher. I was working as a sales in a bank. And I had this problem that my pipeline of leads were a bit messy. I had a lot of leads to contact, but I was a bit lost in who to contact first. I recognize some of them, but there is like approximately 90% of my pipeline. I don't remember them that well. They were leads that were interested once and I kind of lost them. So Cypher is really a tool here to analyze all the data about my leads and to tell me, to push me the right ones I need to reactivate or to contact. I'm going to connect Cypher here. And Cypher is really a tool, uh, designated to analyze all of my pipeline. As you can see, I'm transforming all of my leads of leads, my list of leads in different lists, relevant lists, where I can find my, uh, my leads. For example, I can see Cypher has identified four leads in the re-engaging list. The re-engaging list means that those are leads that were in my CRM for a long time. I didn't contact them and they just made an action on their own. They just re-engaged themselves. Okay. Yeah. Do you have any questions, Sylvain? Uh, yeah, maybe, uh, for instance, this Christophe Gauthier, where could I see the last action you did? I would see that in your Salesforce or? Yeah, it's directly based in my Marketo or in my Salesforce. Like Cypher is plugged to my Marketo or Salesforce. It's analyzing in real time, all the data, each new action will, will have an influence, uh, has have an influence on the, on the score that I'm going to show you after. And of course on the list where they are going to be ordered. So I consider that four people that have just re-engaged themselves, it could be good for me just to send this list to one of my SDR saying, okay, can you check, uh, those four guys, they were in the CRM a long time ago and they just re-engaged themselves. Maybe, maybe the there is a, there is a topic. Um, I'm not going to present all of the list, but two of them, uh, the reactivation list, for example, it's leads that really fit with my ICP, with my persona. And I forgot to contact them back. Uh, so they are really sleeping deep in my CRM. Uh, they fit with my persona, but I kind of forgot them because of the, because of the daily work, blah, blah, blah. I forgot. So those leads all the time, I can open directly the list directly in my Salesforce. So I have nothing to do directly. Cypher is identifying the right leads and pushing it to you in a list saying, okay, why they are in this list. Okay. And here you could, uh, obviously add them to a campaign. If you are connected with, uh, market or HubSpot, uh, marketing, for instance, so that they will be engaged on the marketing side with an email or, uh, an email flow, a nurturing flow, or you could, uh, contact them, send them a message on LinkedIn, whatever, but do your SDR job, right? Yeah. Different scenarios, according to how I am working in my company. For example, if the reactivation list, I prefer to contact them directly with phone. Okay. I'm just sending the list, uh, to my, uh, to my, um, to my sales team. If it's more about a nurturing campaign, which I'm going to show you, uh, right now, actually, I can push them into a campaign. For example, as you can see, Cypher is also ordering your leads that are stuck at one stage of the funnel. Cypher is identifying which content is really working. For example, the webinar of February 24. And it's, uh, Cypher is identifying what content is working at which stage and who hasn't seen this content. So for example, Sandrine, Emily, Dominic, and Michelle, they haven't seen the webinar of February 24, which is a good content at this stage of the funnel. Well, why not send it to them? So this is, this is basically what you said just before I can just now open my list directly here and of course add them to a campaign, to market, to campaign or a sales for a campaign. Oh yeah. So this is the next step because here Cypher tells me not only this guy could be interested, but this is the action you have to do that could fit with their profile, right? Yeah. This is definitely the first version of the next best actions. Okay. I have this person, I don't know what to do with them. Uh, they are stuck at this stage of the funnel for a while. Well, let's see how I could try to reactivate them. So this is basically, uh, the kind of list you will need to, uh, to activate a business in your pipeline. Just a quick question. I see that on all those, all those lines and on all those leads, you have a deal amount and a total of deal amount. So this is something you calculate based on the deals? Absolutely. Absolutely. Most of our clients are tech companies, SAS companies, and most of the time the pricing is, uh, is, uh, is according to, um, to the size of the, of the company or the size of tech team or sales team. So it's quite easy to forecast the, the, the, the deal amount, uh, that, uh, for each, uh, leads. So that, that what we are doing when it's a two specific pricing, uh, we are building it according, uh, to the, with the, with the client. Okay. So here, basically it says that if, uh, if you engage Patricia LaPierre, uh, as, uh, uh, a potential, uh, deal of, of, uh, 3k, three more than 3k. If we can, uh, close her with the article six, which seems to fit her ICP and the step of our lead lifecycle, right? Absolutely. Yeah. The goal is really to optimize your efforts and it's not necessary to, to, to focus, uh, all your attention on a lead. That is a really small deal amount. Cypher is ordering each list, showing you the, the deal amount that represent, uh, that this list represents to help you into, to make the right choice. All right. Between these two, two lists, for example, send the article number six or the lead magnet, uh, lead scoring template, for example, I'm going to maybe, uh, focus more on this list because they have the same, uh, almost the same number of leads, but this one looks, the deal amount looks way bigger. So maybe I can, could prioritize this one first. Okay. Which is, this is, uh, funny because Patricia LaPierre is in two lists, right? I could either send her the article six or the lead scoring template. Okay. Yeah, absolutely. Because Cypher has identified that the lead scoring template and the article number six is good content. And at this stage of the funeral and Patricia LaPierre is stuck at this stage and she hasn't seen neither of them. So maybe it's a good time to, for, for you to explain to us how your AI detect that, uh, at this stage Patricia LaPierre should receive the lead scoring template. How, how do you, uh, do you work to, to, to. Yeah, absolutely. I'm definitely gonna, open the car and show you the motor of, um, of Cypher. So basically we are to know if a lead is a, is a good lead. We are using AI lead scoring. So lead scoring is quite easy. We are focusing on three major data, contact data, who is the person, company data, where is, uh, what is the company where this person is working and, and engagement data, which is the behavioral data, um, that shows how this person, this company are interacting with me and my, and my services. So Cypher as one condition, you need to have enough data, uh, to, to, to install it because of course, if you have, uh, if you don't have any client, for example, or not enough, the, the machine learning model is gonna build an AI lead scoring model that will not be that relevant. So, so one of the, the most important, uh, criteria we need is to have a CRM, to have, to have models of data processing, data collecting, and of course to have at minimum a hundred clients, uh, already signed. And in order to, yeah, in order to determine, uh, the winning side and the losing side, you base your calculation on opportunities, right? Absolutely. It means that the opportunity has to be correctly filled out, which can be a problem sometimes. Yeah, absolutely. Actually, we are helping all of the time. Cypher comes with services. So of course we can, we can help you to, to, to set up well your, your CRM, uh, in this case to of course, maximize the, the, the, the result, the results of the, the solution. Um, well, well, how Cypher is working is that we analyze all the data you have from your current clients. And by it's by understanding how your clients are becoming clients that we can make this prediction on your pipeline of leads. So as you can see, I'm going to open this. I have nothing to do. I have no lead scoring to set up. I don't have to find the criteria. I don't have to guess the number of points I'm going to give. Cypher has been programmed to give you a relevant, uh, lead scoring system going from zero to a hundred and showing you the best criteria for each um, for each data. So as you can see a good contact, the, the, the, the country for good contact is, uh, has been identified by Cypher to be, to be France, the department where the, the, this person is working, the best department is innovation department and the best job title. And I can interact with is a product manager. As you can see, Cypher is always pushing you the second best value also. And the third best value, because of course, we know that you, you don't have only one person as super strict. Of course you are trying to, to focus on new markets. You are trying to focus on new personas. So of course Cypher is identifying what are the best value, the street best value for each criteria to help you. For example, when you are doing the lead generation, um, uh, strategy, when you, uh, try to generate new leads on the, for example, on sales navigator, uh, as you can see the time for the company, for example, best company has a revenue between 10 and 15 million euros. They are tech companies, uh, and the, the size is between 30 and 50. And then the, the part about the engagement, uh, is, is something that always pleased the marketers because of course, this is how you can do a really good, uh, nurturing by knowing at each stage of the funnel, what are the, the best contents that, that, that are really working, that are really efficient, efficient, um, as you can see. Yeah. Tell me. Yeah. For, for the engagement part. So, uh, you will, uh, if I remember correctly, you will use either the campaign, uh, of Salesforce or the campaign of Markito. So campaign and campaign members, right. Or, uh, the activity logs, meaning the, the, the behavioral activity of, uh, that we can find in, in Markito for instance, the fact that open, I mean, in email, you fill out a form or things like that, right? Yeah, absolutely. While, uh, when we are plugging to your Salesforce and to your Markito, all the data of your Markito or the data of the campaigns and the data of your Salesforce are getting analyzed and we are understanding what are really the campaigns that are really working. Is this campaign really efficient? Is, is there a majority of leads going to the next stage after having seen the this campaign? This is basically what we are doing, like kind of statistics. Because last time we talk, I remember I had to do a, a preliminary job that was to bring the data on the person. And now you've progressed and you're able to, to, to analyze behavioral data attached to the contact. Absolutely. Attached to the contact, which is, which is great because we don't have to do this job anymore. Yeah. Yeah, absolutely. Absolutely. We are taking more and more data into account. And the goal is really to help you to take the best decision according to the data you have collected. You have spent money to collect, to collect this data. This is an asset. You definitely need to make it work. So that's great because you here, you have your perfect ICP that maybe you never had any idea of because maybe you said, okay, with, I work with the sales and this is our ICP and the AI will exactly identify your correct ICP. And there may be differences between what you thought and the reality, right? Yeah. I mean, we are, the, the, the clients are really learning things about the ICP. It could be about the job. It could be about the person, the company. Most of the time, it's about the, the behavioral because they don't have the vision that, okay, a good lead. For example, I'm at this stage. A good lead is someone opening at least two email. Okay. That, that means that maybe I should contact them first with two emails and try to, of course, make them visit the website or download this ebook. Maybe I couldn't do maybe an email showing, Hey, check out, check out our website or a second one saying, okay, we just wrote this ebook. You could maybe try, try to check it out. You have the last part is criteria with no impact on score. So let's imagine that we, we see that subscribing to the newsletter has absolutely no impact. And you know, we are spending hours creating newsletters and sometimes definitely it's, it's just a belief that we need a newsletter, but Cypher could honestly tell us with data, that it's relevant or not. Definitely. That's what I wanted to show you because we are at this stage lead and at each stage of the funeral, the engagement criteria are changing because of course a content is, could be good at the lead stage, but maybe not at the MQL stage and so on. So this is what you can see. For example, the, the ebook number two was good at the lead stage, but after at MQL stage, it's more the webinar of February 24 and the article number two and three that are really good. Really working. And as you can see, of course, we can see that for example, the ebook number two is, as I said, useless at this stage. And I think, yeah, for example, as you can see the newsletter at the stage opportunity is useless. That means that maybe this newsletter is something that is really working at the beginning of the funeral. It's, it's useless to push it to a, to an opportunity because at the opportunity stage, they are already, they have already made their, their, their, they have already taken their own decision. They are ready to buy it. They are almost at the end of the funeral. Yeah. Right. In this demo account, of course, according to this data, but could be the, the, what is, what is really relevant is that it's really depending according to your, to your own data. By knowing all of this, of course, you have understood this. Now we can score all of your pipeline and pipeline of leads. So there is another view, that you haven't seen. The first one was all the really operational part. I have all my leads ordered in many lists, different lists. I can attribute them, launch the campaign. So directly, but when I'm coming from, from, from home, the, the, in the morning to work, I have directly my leads fully updated according to all the fresh data I have, and I can address my pipeline of leads. And here I just have all of my pipeline of leads scored with the, the machine learning of Cypher. By, by going down, I can see, of course, the detail of the, of the scoring, but I'm going to go directly on one of them. For example, Michel. Michel has a score of 26, as you, as you can see. The, the fit, the fit with the contact is not really good. It's not really a good, a good part of my ICP. The, the company fit is okay. And the engagement fit is okay also, which is the, the best, but let's see. I have the justification just below. I, I can see that there is in the good company size. There are 50. She visited my website. Okay. It's good. But she's working in a consulting company, which is not really my main target. And she only saw one page of my website. So this is not the best, but by going directly on the next best action, I can see the, the, the next best action that Cypher is pushing it to me that we have seen before in the different lists. And I can see one by one, leads by leads, how I can interact them. This is really, if I want to, to target a specific one in priority, of course, as I told you, everything is sent back in the CRM. So I can open directly my, my, my contact Michel directly in my, in my CRM. And I, and I can see the score of Cypher, the stage, of course, the justification of the scores. We can, I can also put the next best actions, many, many things like that. And I'm really transforming my, my pipeline of this, where I was kind of bit lost. I was like, okay, I know this one. I know that it's going to be soon a good one. This one, I have forgotten all of them here. I get, I'm sending back directly all the data of Cypher and I can see all of my leads with their score. And so I can prioritize in this way also. Okay. So when we, when, when you do the project, when you start a new customer, you will integrate Cypher with your Salesforce and the marketing automation solution. You will create, you will ask your customer to create fields on the contact and lead, I assume. We are creating them directly because when we are connecting, we just are like plugging with an API and we are automatically generating them in your CRM. You just need to make them appear on your screen. The properties have been created, but the need to be visible on the, on the dashboard. And so are the list that you were talking about? Absolutely. Everything. Yeah. I forgot to say that. Of course, Cypher is updating every hours. So of course, by sending a new market to campaign, for example, a new nurturing market to campaign, you can see directly the scores moving one hour after you can see the people going, going out of the list because of course, now that I have received this article, okay, I'm good. I just received it. So I'm leaving this list and, but you can see also new leads coming inside the list. And of course, one of the goal of our clients is always to try to, to make the different list as empty as possible. Of course. All right. Yeah. It's a great help. You told me that, yeah, that every hour you sent and you, you did that, but the model is refreshed every day. Right. Which is great because in real life, I think the scoring models are refreshed every six months or every year. If, if there are refreshed at all. Yeah. The, the good students, the good companies that I was interviewing, the good students, they were really refreshing every three months, which is really good. Good. Good thing. But most of the time, yeah, you are set up in early scoring and you are never reworking on it. Yeah. When you say we refresh, uh, I'm not sure that the customer refresh actually do the job of taking all the opportunities, all the contact data, all the company data, all the behavior to see if there are patterns to re I'll say those are the things that we have worn in the last three months. Here are all the patterns we have, and we changed the scoring according to that. It's more or less. I think that's, yeah. In the newsletter, we should add two points. I think the video, we should remove two points. It's more like seeing like that. So it's not so much data-based. It's always like that. A manual scoring is no matter what always biased, uh, always biased. And of course the best way is to give all this data, all this kind of wild data to an algorithm, to, to, to ask him that is not biased, of course, to ask him, okay, try to find some pattern, try to find what, what's my best, my best person. What's the best content at each stage. And this is how. We can really offer you a really relevant, uh, scoring system. And you told me you already have some results to prove that this is, this is actually working, right? Sure. Sure. Most of the time, one of the good statistic that I have, uh, I have seen, uh, while I was, uh, building Cypher at the, at the beginning is that from a company that is, that has no lead scoring, even, even a manual one and a company that has, that has one. That has set up a normal one, uh, even a manual one. Sometimes I was seeing an increasing of 20% of the conversion rate between MQL to SQL, because most of the time, this is how the, this is how the, this is why you, you, uh, you are set up in a lead scoring because you are, you have a lot of MQL and you have made by the marketing and you have sales and they want, of course you want to send to the sales only the good ones, because of course, if you are sending the bad ones, the sales is going to be okay. You have, you're sending me super terrible quality, low quality leads. Can you please try to do something? So this is why you are set up in a lead scoring, trying to include all the data. So from a, from a manual one to a normal, to, to nothing, sometimes I have seen in the best case, of course, 20% increasing, uh, in the conversion rate between MQL to SQL. And we have seen from a manual one to a, to a, to Cypher, to a Cypher one, we have seen 10% more increasing in, uh, at, at our clients, uh, at our clients, uh, from this part. We also, it's some, sometimes it's quite hard. Also, it's, it's depending on the, the, the clients, of course, but what we are also tracking a lot is the different leads that were in the list that I, I have showed you, uh, just, just before, and that are finally becoming clients. So we are really focusing on, uh, this part and saying after a few months, okay, uh, there was this four people, uh, in this list, in this list, two of them have left as, have left the list and they are already moving to the next stage. And for example, one of them has already been signed. So you have reengaged leads that you would have never reengaged without us, because of course, those are leads that you were not focusing on because you are only focusing on the new, new pipeline. Uh, you were not focusing on the, the sleeping leads. So of course we are more focusing into this and you are, uh, receiving daily reports every week, uh, saying, okay, in this lead, in this list, there were that much, uh, leads, uh, few of them have been converted into clients and few of them have been, uh, moved to the next stage. They have progressed to the next stage. Yeah. And here we, I think we, we, we touched the future of marketing, uh, marketing will be in charge of creating those, uh, wonderful experiences such as the, you know, the videos, the blog articles, the eBooks, the, the quiz, the infographics. But I think AI will be in charge of deciding what's the next action for this prospect, because there are so many possibilities that you, we humans cannot create all the customer journeys that match all the possible configurations. It's not humanly possible, you know? And yeah, that, that I love that because that's, that's what the future looks like, you know? Yeah, definitely. Let's focus on really high value tasks, high value job that human can do and let the analyze of the, all the, the, the data to, to, to an algorithm. Because of course the marketer is not really, or the sales manager is not really someone that are analysts that is ready to, to analyze a huge volume of data. Most of our clients have between like 50 K to 80 K contact. Uh, so it's quite a lot. Most of the time you only remember 10% of it. Yeah. Yeah. So you have half your database, which is sleeping somewhere and never being taken care of. And thanks to, to Cypher, you can identify within this, uh, I don't know, 40 K contact, which one are the one that merits your attention. Sure. Sure. Definitely. This is really the main goal of our, of the company is like re-engage the leads that are sleeping in your, in your CRM instead of always, okay. Generating new leads, generating new leads, scrapping another database of a sales navigator. I don't know, uh, try to, okay. The business is already inside. Of course. And when you start a project, how long do you need? Do you, AI needs to learn everything and starting giving you a real advice? So, so Cypher is, is, is, is, is a, is a solution made for kind of big companies. What I mean is like between, uh, let's say more than, more than 50 employees. Most of the time, this is like the main target. Yeah. For me, it's a bit, uh, we don't have enough data, I think. Yeah. Yeah. Yeah. Yeah. Yeah. Well, we have tried, we have met, make some, some things, but of course, yeah, you, you need a huge volume of data. You need to be facing like the challenge of, okay, I have a lot of leads in my CRM. I don't know who to contact first. I have sales. They need to be nurtured with leads. They need to be fed. Okay. I'm just plugging Cypher and I'm going to check, uh, what, what is going to push me. Um, we are, when we are installing Cypher, set up in Cypher, um, we need to have three weeks of setup. Uh, and after this, uh, this phase, you are fully, uh, autonomous on, on the, the, the solution, because of course it's not really changing your daily life. You have all the data already in your CRM and you are just taking a better decision with the data inside. And when you face during the setup phase, uh, bad quality data, missing data, how can you help? Yeah. So most of the, most of the time we are working with partners that are data specialists. They are, uh, Salesforce and CRM experts. So we can, uh, had some, uh, features, some process of, to collect data in a better way to collect data, to, to, to make it analyzed by Cypher. And I haven't faced the problem where the CRM was too messy to, um, to, to in low, low in data to, to not, uh, install Cypher. Most of the time, I know that people, they don't want to show their data. The CRM, it's like almost getting naked in front of strangers. Uh, so this is really the problem, but trust me, like we have seen some really messy CRM and most of the time there is always good things to do to build with even. Yeah. What kind of data can you enrich? Uh, if something is missing? Yeah. Most of the time we are working with open source database of companies. So when you have, when you have, uh, the number of a company with what we call the, the, uh, the number of the, the, the company, we can, we can say, okay, this is not only this company, this is a company of 50 employees that is based in Paris that has raised funds. That is a profitable that we can enrich all of this kind of data. So it's helping you into, as you, as we said just before about the personnel, we are helping you into refining again and again, your personal training to, to make you learn more things about your, your target. So it's three weeks to, to the setup so that we can have a ready decision. Yeah. There's a, when you create, for instance, a new graphics, a new ebook, a new article, um, how, how can AI learn, uh, which, which ICP is gonna, well, how, how the learning curve is happening. If AI is deciding everything at the end, how can we introduce a new element in the game? Definitely. So let's imagine that I'm okay. Cypher, I have connected Cypher and, uh, today I've decided to, to launch a new ebook, a new one that we just wrote. Um, and, um, in this case, Cypher, as we said, every day, Cypher is re analyzing all the data again, every day at three in the morning, uh, at 3 AM. Um, in this case, not directly because of course Cypher will see during the few days, he will see, okay, there is a new ebook. Let's see how people are interacting with it. So we will see Cypher will analyze all the people that will check at this ebook will download this ebook. And of course, Cypher will see if the majority of them are moving to the next stage. So, and if there is a majority of them, let's say that this is an ebook that I want to use for my MQL. For example, I want to reactivate my MQL. I have made this ebook for them. That was, that was my target. So I'm going to send, send it to, to all of my MQL that are stuck at this stage. Uh, it's quite easy with Cypher. I have a list. So directly I can do some test and, um, I'm sending to them. And after a few days, I'm going to check if a part of them, a majority of them has moved to the stage SQL. If yes, if, if there is a majority of them that have checked the ebook before moving to the next stage, in this case, after a few days, Cypher is going to say, okay, now I've moved to the next stage. Modified the, the lead scoring system. So this part that I've shown you, and for example, I'm going to, I'm going to see here the new ebook, uh, appearing here saying, okay, this ebook, this content, I, I, I think that this is a, this content is really relevant. So I'm going to add it there. It's really according to your sales cycle. Of course, if you are taking a year to sign a client, it can take more time. If it's a few months, of course, after a few weeks, you're going to see it. It appeared here. Yeah. So do we need to wait for a, an opportunity one to learn that or just the next phase? The next stage. Okay. Yeah. I was saying it a bit fast, uh, at the beginning of the interview, but we are, we are providing an AI lead scoring system at each stage of the funnel, which, which actually nobody is doing this, but I don't know. I don't know why, because it's quite useless to bring an AI lead scoring, one lead scoring for all of your pipeline. Of course, your leads at the beginning of the pipeline are not going to be scored well, and the opportunity are going to be super scored, but you don't care because of course, it's not interesting for you to know that this opportunity is at 88. Yeah. What you want to know is which MQL has a good score and is ready to move to the next stage because the main efforts are going to be at the beginning of the funnel. So this is why, what we are doing. Yeah. Oh, yeah. So I, yeah, it means that the marketing has still a role to decide, uh, which content to create, how we promote this content at the beginning when AI has not learned yet, uh, if the ebook is relevant or not, but after a while you can, uh, use a cipher data to say, okay, we created this ebook for the MQL and we see it's working great. So let's, uh, let's industrialize that so that we, uh, send this ebook to everyone. You have this, uh, from Salesforce, right? Absolutely. Yeah. Uh, yeah, yeah. Uh, Einstein from Salesforce. Um, it's a super complicated solution that requires you to hire at a minimum, a data scientist or data analyst, uh, inside to code some part of the AI lead scoring. Well, this is not the premise. No, I was thinking that, okay, I'm buying Einstein. Okay. Do the same as cipher is doing. We are just analyze all my data and push me the right lead scoring system. No, you, you really need to include a different data and it's including a part of the codes, uh, that you, you can't avoid escape. Uh, so, so yeah, the, the pricing is, we have, we have just set up fees for the three weeks and after it's just a pricing according to the data you have and it starts at $700. So how many, how much? $700 or euros. $700 per month. No, I'm right. Yeah. So for, for less than 10 K a year, you have the full pack of cipher, right? Yeah. Yeah. Yeah. Yeah, definitely. Yeah. Which is, which is great. If I compare to, uh, trying to, uh, build up my old model and, uh, training it and doing the integration and which I do that internally, that could take months and be not of the best quality. Uh, and it, the, the IT department may not be, uh, push that as much as possible. It's a priority. So it is great that under three weeks I can have that working and giving me a good decision for myself and the marketing that that's a great. Yeah, definitely. Most of the solution, the current solution are not providing a precise, a precise model like, like us. And if you try to build it internally, you need to have a good marketer because of course he needs to know he needs to do statistics. He needs to really analyze the data, uh, the past data to try to, to, to build this model. And of course we already said it, but it's going to be biased anyway, uh, because it's not, uh, because it's not based on the, on the code. And one thing you told me that it's great for, especially for European clients, customers is that, uh, so yeah, we connect, uh, cipher to our solutions. So all that goes to cipher on your servers, which are on in Europe, right? Yeah. Yeah. That doesn't exit cipher. We, you don't push that to chat GPT or any other, uh, AI model. Yeah, absolutely. Yeah. I would like to precise that we are using machine learning. This is a family. This is a part of the AI. We, everyone knows AI now since two years because of chat GPT, of course. Um, but we are not doing LLM. LLM is to, to, to, to, to build, uh, to write things, to build videos, to, to do things when you, when you want to, to analyze huge volume of data. Machine learning is, is the best, uh, family of AI that you can use. We are using models that have like more than 30 years. They are more than 30 years old. Um, but they are quite, quite efficient. They are used everywhere. So this is basically what we are doing. Our server are based in Germany. Now we just moved them before they were in the UK. So no Amazon, we are, um, only analyzing, the data you have already collected. So you collect the data and my job as cipher is only to say, okay, this guy, this guy is interesting. You should collect, contact him first. We, we are only analyzing the data. And of course, everything is, is closed. Um, nothing is sold, of course, to other company, company or whatever. And, uh, nothing of course is, uh, sent, uh, into chat GPT because there is no, the first it's forbidden and there is no, no real reason. Of course. Okay. Well, it's a great, uh, value for, uh, any sales or marketing team having, uh, a CRM and a marketing automation. And, uh, I think in the very near future, it will be a must have for any, any company. I'm convinced. Um, any, any, any word, Paul, regarding that? No, I would be, I would be happy to talk with you. Do not hesitate to add me on LinkedIn, of course. And, uh, and to ask for a demo, it would be good. We are connecting to all the common tools of the market. So today it was about Salesforce market though, uh, because of course, uh, Sylvain is an expert in, uh, with these two tools. But of course we are also doing HubSpot, PipeDrive, uh, many others. Uh, yeah. And that's all. You can connect to any CRM tool of the, of the market. Yeah, absolutely. Okay. Yeah. Maybe, maybe some exotic CRM, uh, I'm not doing it, but you know, the most common ones. Yeah. No problem. All right. Thank you very much for your time. And, uh, we'll keep you posted for the, for the next evolution of, uh, of Cypher. Thank you. Thank you, Paul. Bye. Bye.