Modicus Prime Safeguards Drug Manufacturing

The Harry Glorikian Show

Final Transcript

Harry Glorikian: Hello. Welcome to The Harry Glorikian Show, where we dive into the tech-driven future of healthcare.

Quality control is one of those things that nobody cares about until something goes wrong.

That’s doubly true in the pharmaceutical business.

Episodes like cross-contamination in a drug factory can shut down a production line and create instant shortages of important medicines.

And if a contaminated medicines ever does get shipped out to clinics or stores, people’s lives can be at stake.

So drug makers are usually pretty receptive toward any new technology that can help them detect manufacturing problems before they get out of hand.

That’s the market opening that Taylor Chartier says she saw back in 2020, during the coronavirus pandemic.

You probably remember the stories about the Baltimore company Emergent BioSolutions, which was manufacturing vaccines for Johnson & Johnson and AstraZeneca and had to throw out millions of doses of both vaccines due to suspected cross-contamination.

Chartier saw those stories too. And she thought, there has to be a better way.
So she started her own company.

And today her startup Modicus Prime is partnering with top pharma companies to use new AI capabilities to catch drug manufacturing problems faster.

She says the company’s machine vision software can analyze data from existing cameras and other hardware on a production line and identify a problem with 0.6 seconds of capturing the image.

That’s fast enough to shut down a line and fix a problem immediately, without having to throw out an entire batch of pills or proteins, the way most drug makers do today.

Obviously that could mean millions of dollars in savings.

And ultimately, AI-enhanced forms of quality control promise to lower costs across the whole healthcare industry.

So I wanted to have Chartier on the show to talk it all through.

I especially related to her comments about why she’s a fan of mixed martial arts and how starting a healthcare company is a little bit like ultimate fighting.

For that, you’ll have to listen all the way to the end of the interview.

We started out by talking about the dimensions of the quality control challenge in the drug business.

Harry Glorikian: Taylor, welcome to the show.

Taylor Chartier: Thank you.

Harry Glorikian: So, you know, it’s great to have you on the show. I mean, I’ve known about the company for quite some time. You know, even before you and I had a chance to meet in Harvard Square. And, you know, when you explain the technology to me, I’m like, yeah, yeah, I totally get it. I know exactly what you do. You know, it’s I think it’s a good thing because it just clicked very quickly for me. But I and I know that the main selling point for Modicus Prime’s AI service is something called MPvision. And, you know, that’s something that helps companies in research or manufacturing drugs maintain higher quality control. But I wanted to I was thinking about it. Before we jump into MPvision specifically, can you help listeners, you know, maybe explain to them what is the main quality control challenge, you know, that that drug industries are sort of facing? What are the different ways processes break down or production lines can become contaminated? How do these problems come up? You know, are they the result of human error? Why are they so hard to detect these problems? I mean, maybe a primer on that before we go into this, because your technology is sort of pointed right at some of these issues.

Taylor Chartier: That’s a great idea, Harry. Thank you so much. I’d be happy to provide some context on the pharma industry in general, especially when it comes to quality control. It’s a bit of a niche industry when you talk about biopharma. So happy to shed some light on the topic. Definitely in terms of the public eye and what was really highlighted during the pandemic, Covid-19 just really opened up Pandora’s box. There were a lot of reports and the media covered cross-contaminations with, as we are all familiar with, with Johnson & Johnson’s product, with AstraZeneca seeing hundreds of millions of doses being discarded due to contaminations, it was very disheartening to the public. But what’s really interesting is that contaminations are actually systemic to the industry. This was not something that just happened in during the pandemic, as we know. And in fact, every year the pharmaceutical company on average will have a drug product quality failure that occurs. It’s a $50 billion resource. Unfortunately, it looks like every year these these companies are continuing to to struggle to maintain optimal quality control. And we’re looking at it in terms of when we talk about contaminants, as you had said. A lot of these contaminants, about a quarter of them actually come what they call are particulates. And these can come from a variety of sources. They can come from the equipment, they can come from the particular vials. For example, if it’s delamination, they can come from glass particulates that come from the vials of, for example, a vaccine. They can also be introduced inadvertently from the the technicians. Could be pieces of hair, pieces of rubber. There’s a vast variety of of contaminations that that that can be introduced in the during the drug production.

Harry Glorikian: So, you know, I think for people in this area and when I mean local, I mean the local Boston area, I’m sure that most people who’ve been in the industry remember back to 2009, when a Genzyme factory here in the Boston area that made protein drugs discovered, like this viral contaminant in their bioreactors, and they had to shut down the whole plant. It took ten months to identify the virus, track down the source, disinfect the whole place, and then restart everything. Is that the kind of problem Modicus Prime can help prevent?

Taylor Chartier: Yes, exactly. Genzyme has been through it. They all have. And it’s just it really points to the complication of creating what has been called recombinant proteins. These are not your typical Advil or, you know, some of your other small drug molecule drugs. These are very complicated, when a host organism is grown in a bioreactor, as you were referring to. And then the manufacturing process that takes place is really just to isolate and purify this protein, which is this active ingredient from these cells. Oftentimes those can be Chinese hamster ovary cells, and you just are extracting this protein from them. It’s massively complicated with many purification stages. And exactly to your point, this is exactly what mpVision can help address, either, whether it’s at the bioreactor stage during the fermentation process or at any stage subsequently during the purification steps, when you’re going through chromatography columns and your different types of viral inactivation, all those stages, there’s quality points at every step. And and that’s exactly the challenge that that pharmaceutical industry faces.

Harry Glorikian: So I want to take a step back, just one step, right. And just have you give us a little bit on your career path. Right? I mean, how did you decide to build a career around quality control and drug making? Not something that somebody just slides right into as far as I know. But, you know, maybe start with your training. I mean, as far as I know, you have an MS in chemical engineering and a PhD in AI applications in systems biomedicine. And to be honest, I didn’t even know there was such a degree. And and you were at Bayer for a while, I mean, one. Tell us walk us through like you know, how you got from where you were to where you are now.

Taylor Chartier: Oh, I’m happy to do so. And just touching back on, on that, that you were referring to, I was actually at the University of Luxembourg pursuing a PhD when Modicus Prime was born. The entire school system and many others in Europe also, they actually shut down temporarily. And what better things are there to do with your spare time than to start a company? It was pharma was was working full speed ahead and I had actually, formerly I was working in pharma. More on the data science side after University of Rochester with my chemical engineering degree. I popped over to Shire in Boston prior to its acquisition by Takeda, and just had the opportunity to really support various biologics license applications and perform root cause analysis for quality issues. And I would apply AI and ML in those capacities to really help optimize pharmaceutical production. So when I saw all the things that were transpiring with pharma during that time, and there was that pause in my life, I thought, well, I’m just going to reach out to some of my former colleagues and see, hey, how can we help? I know exactly the types of algorithms that can be used to address some of these issues, and it was one of those classic PhD dropout stories where the opportunity was there, the timing was right, and our first customer was actually a top ten pharma. We incorporated, jumped right in and we’re just had an amazing partnership from the start.

Harry Glorikian: So maybe now’s a good time to sort of maybe dig in a little bit into the inner workings of mpVision. I mean, personally, I don’t need to be sold on the idea that computer vision and AI can transform the drug industry process. But I’d like, you know, if you could help me sell the listeners on that idea because it’s something that, you know, coming from pathology, visualizing, diagnosing, identifying. I mean, this is this is a very common sort of thing. But if you can do it in process and find something very quickly, it can sort of help you triage and manage the process. But, you know, help us understand, you know, what happened in the last few years on both the vision side and AI side to create these new capabilities. You know, what does vision do? You know, that sort of thing?

Taylor Chartier: You mentioned pathology. Pathology. They have just been a North Star. Taking a look at medical devices already, there are over 500 medical devices that have been approved by the FDA. Just really tremendous progress has been made in the space. And it makes sense specifically when you look at pathology, about three quarters of all of these medical devices are for specifically imaging. And if you look at hospital data, 80 to 90% of that are images. So computer vision is is definitely the most mature AI type of application. And pharma has done a really great job of leveraging a lot of the guidances for what they call AI software as a medical device. They’ve taken a lot of that around 2021 and and kind of brought that over to now the manufacturing side. And you’re looking at the same type of of outputs that you look for in pathology. It’s very similar in terms of what you’re looking for. You’re looking for, for example, anomalies. So instead of looking for an anomaly like a tumor, you’re looking for an anomaly of, well, is this a heat signature that shows there was a temperature spike during production? Or is there some sort of morphological change that could indicate there was some shear stress in the bioreactor? These are the types of visual things that you can see and characterize in, for example, with vision.

Taylor Chartier: So we’re really excited to see a lot of different pieces come into play with the pharmaceutical production industry. And we’ve seen just last year there are other organizations that have helped further translate FDA guidances into actions that the industry can actually implement. So GAMP, good automated manufacturing practice. They have their second edition published last year, and it really shows in a systematic way how to take AI computerized systems and validate them for production. That was a huge question because compared to traditional computerized systems in pharma, they are not data driven. So now we really have all the pieces in place and we’re seeing actually in production, we’re seeing visual inspections, we’re seeing chromatography applications. These are now, these are now reality. And though not everybody is familiar with it, the path is is really clear for more AI technologies like mpVision to be integrated in pharma.

Harry Glorikian: So tell us how this works. I mean is it you know do you extract a sample and put it under a microscope? Is there a camera that’s in line in the process? Do they have to train their platform? I mean, just give us a visual for how this is how you would put this into play.

Taylor Chartier: Yes. So you can consider vision as a software layer that you can add to any preexisting imaging hardware at any unit operation in pharma. So we’ve already talked about how many images there are in hospital systems, pharma. The same thing we’re talking about, you mentioned microscopes, microscopes, image flow, cytometers, micro flow imaging devices. There are just a myriad of of technologies that are already being used for quality purposes in pharma. So there really is no additional exchange, no upgrade of hardware that needs to take place. The software directly integrates via our API with any of these preexisting hardwares. And then once these images are actually uploaded to our system, we are cloud based, though we do also have on-prem offerings. Once these images are essentially collected, it’s enters then our workflow. We have a platform that’s end to end, and it really helps whoever the end user is. If it’s a researcher or a domain expert, a technician, they’re able to if they know the product, they’re able to use our platform to upload these images and then teach the AI what specific type of quality characteristics that they want that AI to detect.

Taylor Chartier: We were talking earlier about how it could be a shear stress or heat stress. It could be an anomaly. There could be some type of, for example, an air bubble if they wanted to detect or silicone oil particle, anything that they think is a significant to the quality of their product. They’re able to teach that AI what it is. And just with a click of a button, the AI model, it’s all encapsulated. It starts learning, what are these features? And traditionally there’s this what’s called feature engineering. And that’s when someone might use mechanistic modeling. And they come in with different parameters and decide, okay, you know, this feature is significant and will help differentiate what’s a contaminant, what isn’t. That’s the beauty of AI. It will do that all for the end users. And at the end of the training period, you end up with an AI model that you can roll out for real time quality control that already has learned everything that you have taught it to from that data set you introduced it to.

Harry Glorikian: But I would assume that, you know, once you’ve seen a certain type of contaminant or a bug that you need to identify, you probably have a library that people can access and utilize in their every day so they don’t have to train again if they don’t want to.

Taylor Chartier: Yes, that is one of the huge value adds of our technology. Like you said, a library. That is exactly the desired output for for many of our end users. They have a very unique portfolio. So in each of their the drugs that they’re manufacturing, they may have certain contaminants or certain quality characteristics unique to each one. So they can create their own library. But one of the another one of our benefits is we do not take their data and save models that we then distribute to other pharma companies. We don’t do that. It is all confidential to them. And if they want to create their own unique library, that is their own competitive edge, that that library is theirs and we’re happy to support them.

Harry Glorikian: So you mentioned something because I was under the impression that the business was a a SaaS style business, or for those listening software as a service, but you mentioned that you also have an on-prem, which means they would be able to have a, you know, process where they would be able to download this or have this specifically at their site, as opposed to being up in the cloud. Is there a big difference for different users or is one more powerful than the other?

Taylor Chartier: We do see there are differences in terms of of scalability, but often times it’s actually preferable on the pharmaceutical side to to actually have this offering on-prem, because in many of these cases, they’re actually the ones who are in control of of the compute GPUs is, as we all know, that’s a huge requirement of this technology. And in times where a pharmaceutical company might not know beforehand how much image volume that they’re going to be processing, they may want to have control over the amount of GPUs or the amount of images that they process and have a better understanding of that. So in terms of our business model, it’s oftentimes where if the pharmaceutical company offers to pay for the compute, they’ll have asked for an on-prem, and then they take care of that on their side. And then we just provide the support for the infrastructure.

Harry Glorikian: Yeah, I can imagine the smaller ones are probably willing to maybe do go to the cloud and then the bigger ones might want it, might want more control over the process.

Taylor Chartier: Yes, yes, exactly.

Harry Glorikian: So you mentioned some of the defects and and contaminations that the algorithm can be trained to look for. But I guess here’s a question. So you’re looking at this. How fast does it detect the anomalies. Does it does it happen fast enough to shut down a continuous process before a whole batch of medication gets ruined? I mean, these are the sorts of questions I’m thinking of, like that value add that someone gets from having real time feedback.

Taylor Chartier: Absolutely. This is absolutely the huge output from using this technology and what we call a GMP setting, good manufacturing process setting. So in manufacturing, the typical workflow now is if any type of quality characteristic that may be negative is determined, the entire batch has to be tossed. The FDA will just say get rid of it. You know, that’s it. And that’s often times because in during visual inspection, for example, as one approach to quality control, during your final fill and finish, you take a statistical approach where a sample is taken and it’s supposed to represent an entire population of all of the vials, for example. So if that vial is is contaminated, then it can be assumed that it is representing the rest of the population right now, that is obviously very limiting. So when you have an, you know, compared to that approach, a real time system that’s used to detect different types of contaminants in real time, that is where the real time processing is absolutely crucial. That’s where the compute comes in. And and exactly with our technology, we’re able to detect things within 0.6 seconds so that in the event that there is during a batch, there is some sort of of contaminant detected, the end users can actually be able to stop and isolate where that batch is, has an issue, and then release the rest of the the batch to the public, saving them millions of dollars and also saving the public health as well.

Harry Glorikian: So maybe now’s the time to sort of step back. And I was wondering sort of the type of AI that Modicus Prime is using. I mean, especially these days, right now, all the hype is around generative models and, you know, more powerful and more capabilities and ingesting more data. And, you know, what is it? You know, the overlords are going to come and get us. But, you know, I would almost call that sort of adaptive AI, but you guys are using a different kind of AI. I mean, I think you’re using what’s called static AI versus what I just referred to as adaptive AI. Can you you know which one is appropriate in sort of the high stakes applications like you’re talking about in drug manufacturing and why?

Taylor Chartier: Yes. You’re exactly distinguishing the two various buckets of AI that we see today. And for sure, any time that you’re in a regulated environment, static modeling, that is exactly the way to go. We’ve seen tremendous success with static models. And and just to kind of describe what that means by static, the building of the model is still a very iterative process. And so by that we mean that when you’re introducing this model, for example, operationally during production, you are continuously monitoring it, the inputs, the outputs. And if there is for example, a new type of anomaly that’s detected, you are then able to state, well, we haven’t seen this anomaly before, and our static model detected it as an anomaly, but it wasn’t able to characterize what it is. We want it to know what this is. We want it to understand the next time we see this foreign object, we want it to be able to detect and tell us what it is. So then as part of this, this static model operational process, we can actually introduce then that foreign particulate into a training data set and then have that model learn it so that here’s where the iteration comes in. And then you redeploy then that static model. So now it knows exactly what that new particulate is. And if it comes again to detect it. And you can do this every time you know something changes like the data input changes. But once it’s been deployed that performance it is expected. It’s anticipated. You’re able to monitor it. As part of mpVision we actually have an AI model calibration module, and you’re able to actually visualize and see and confirm that, yes, the model performance in the architecture has not changed. It is the same. And yes, we can improve the performance. But for now, you know, this is this is what we’re deploying. And we can be confident and and trust the outputs.

Harry Glorikian: So I don’t know if you have maybe a case study you can share, but I’m sort of curious about like, you know, what happens when someone adopts this process, you know, can it speed up manufacturing? Can it reduce costs? Can it help with regulatory compliance? Is there is there some I don’t want to say a canned case study that I’m sure I’m not the only person that’s ever asked this, but that, you know, that asks this sort of question of like in a real world example, you know, how did this get put in place? And what are the what are the levers it pulled to make a big difference?

Taylor Chartier: Yeah. I would be very happy to share with you many of our implementations and use cases, as we understand, you know, this is very confidential and we do respect the privacy of our end users and our customers. But I will share with you very specifically the different use cases. We have been talking today about contaminations, and that is absolutely a massive issue. But when we are referring to the overall safety, integrity, purity, efficacy, the PK of drugs, there are many other facets of quality control. Right. And mpVision actually addresses those as well. Our end users have been very creative in not you know, there are so many different images as we’ve shared and so many different spaces. They’ve actually been almost educating us as well what unmet needs and quality that they can implement in mpVision to address. So it’s very specifically there are a couple of main verticals that we’ve seen. We already talked about the contamination of the quality control, where we’ve actually been able to demonstrate that it would be a 60 x cost savings with continuous processing with vision, and we also have over 99.9% accuracy that we’ve demonstrated to one of our top ten pharma customers in detecting particular contaminants that are unique to them. So that has been, that’s definitely a massive use case. But what we’ve also seen is if we shift back, we were talking about bioreactors, right. So in the very beginning where you have the proliferation of these host cells, this is a very subjective process. It’s like a cauldron, right. And you’re tossing in all these ingredients and you’re growing, you’re like a gardener in a way. You’re growing these cells that have this drug product inside. Well, the cells, if you want to think about it, sometimes they’re happy and sometimes they’re not.

Harry Glorikian: I was going to say happy cells. It’s all about the happy cells.

Taylor Chartier: Yes, the viable ones, the viability. That’s actually one of the KPIs, the key performance indicators that you look for. It’s like a quality check to measure the happiness of these cells. So if you do not take care of the cells and their terms for that, passaging the cells and helping them to grow to the next stage to help them create more of your drug. If you don’t do that correctly after 45 days, for example, of processing some of our customers, they will experience what they call batch failures. They’ll lose everything. Everything. This is not at the end of the final product where you have vials. This is early on, where again, an amazing amount of resources have been put into this process and the cells die and or, you know, there’s there’s some types of contamination. So what we’ve seen is there are very particular morphological features that can be used and monitored to identify the exact time when the end users can decide whether to passage the cells or not. You can think of it like a little a short time window to avoid these batch failures from these, these cells early on. So that is a huge quality issue that is being used now for our end users to prevent these batch failures. That’s one group.

Taylor Chartier: We also mentioned another vertical. We were talking about having the the cells produce more drug. So that’s referred to as titer oftentimes with the proteins specifically. And you really want to make sure that that titer or that drug you’re producing from these host cell organisms is the highest that you can make it. And there are certain types of morphological features that can be tied to when that titer is the highest. And that way you meet those those quality expectations of how much drug that you have when you when you’re at the final production line. So we’ve actually been able to implement there for understanding how we can be consistent with when the cells are happy and when they’re producing the maximum amount of drug.

Taylor Chartier: And then I’ll say the last vertical, this one has been absolutely, really, really pretty, really tremendous from our side because it’s in the cell and gene therapy space. We’ve been talking about recombinant proteins. Historically, this is how traditionally a lot of these different hormones and enzyme therapies that we are, the scientific community is providing to patients. This is, that’s a typical methodology. But selling gene therapies now we’re moving away from the proteins that have been, you know, engineered. Now we’re moving into what’s called cellular therapeutics. And in this space this is where mpVision actually has what we are discovering the highest unmet need and the AAVs. So vectors, adeno associated viruses, it’s a new methodology. So instead of treating with a protein like for example insulin you’re treating patients, now we’re moving into looking at cells. And these can be if you’re looking at car-T’s for example, you’re looking at your own T cells, if it’s autologous or other individuals T cells if it’s allogeneic. So now it’s a different modality. But this is just another vertical that we’re seeing mpVision implemented in.

Harry Glorikian: That’s great. I mean it seems like the opportunities are are fairly endless. Although as a startup you got to pick one at least and turn the crank as many times as you can for, for revenue standpoint. But so about a year ago, I was talking to Erwin, who’s the CEO of ELabNext, all about Digital Lab network software and. I think you guys have a partnership with them. So can you talk about how mpVision integrates within eLabNext software and then what do customers do with it?

Taylor Chartier: Yes, yes. We’ve been talking about the pharmaceutical space. And within pharma manufacturing there’s also your drug development, your drug research. You know a little upstream of that manufacturing. All of it is tied together. Every stage there should always be that element of of quality, repeatability, transparency. And so at every stage we see opportunities for AI to be used when it comes to more of the laboratory research. Which eLabNext is is fantastic. For they have a electronic laboratory notebook offering that we are integrated into currently. So globally ELabNext is a fantastic partner. We are one of their add ons in their marketplace. They have many a myriad of other add ons that they also offer as well. And we’re one of their offerings. So all of these labs who are experiencing quality issues on their side, they’re actually able to leverage us. We’ve seen actually the numbers just taking a step back. It’s been reported that $28 billion is lost each year due to experiments that are either irreproducible or have errors in them. It’s a really difficult thing to do research number one, and then to do it again. Number two is even harder. And so a lot of times that unfortunately it can lead to some some errors and waste of time. We’ve seen from our own actual customers that they’ve reported that for some of these softwares they’re using for their experiments and life sciences, they unfortunately say that it’s about an hour and a half of just simple, you know, software recalibration and like traditional things that sometimes take so much manual time and sometimes they introduce human error there. Those are things that we help completely eradicate just due to the AI aspect of our in the generalizability of our offering. So they’re early, they’re upstream. A lot of them are working on drugs, but with lab customers, they’re they’re very wide across life sciences to agriculture. We’ve seen, you know, a tremendous amount of use cases in other areas as well with them.

Harry Glorikian: So I’m going to take a little detour because I noticed on the company blog there’s a post called Fighters and Founders, where you compare startups to ultimate fighting or mixed martial arts. And I feel like that sometimes myself in this, in this world. But but there’s even a snapshot of you and your engineering director with Valentina Shevchenko. Thank you. I’m like, I’m going to I’m the poor. I’m. I apologize, but but she’s a former world flyweight champion in ultimate fighting, if I’m not mistaken. Um, what was the message in that post? I mean, what were you trying to get across? And then you’ll forgive me, but I don’t remember you being a UFC or MMA fighter yourself. I don’t know if you’re a fan, but maybe you can give us a little bit of insight on on that post and, and where you were trying to go with it.

Taylor Chartier: Yes, I’ll be happy to. So related to the personal side for UFC, I am a huge fan. I’m a tremendous fan of UFC. So the director of engineering who you see there, David, so his brother is actually a professional MMA fighter and he is currently actually in Dublin training with Conor McGregor’s gym. He’s training. So we we’re huge fans of MMA in UFC in general, but that the specific reasoning behind that post, I just I have so much tremendous respect for a lot of these fighters that you see every, you know, every Saturday or so there’s a fight night or, you know, there’s a pay per view and some of the the champions when you hear what they say after the fights, one of them, Sean Strickland, he’s the most recent middleweight champion. He beat Adesanya. It was a huge upset. One of the things that he said before was look, when opportunity comes, it’s usually comes knocking at the worst possible time. You’re not ready. You don’t think you have everything in place. You know, you don’t want to put in all the effort because it’s two weeks notice, for example. And he says, you just have to do it. You just have to go. You just this is what you worked for this whole your whole life.

Taylor Chartier: You just have to do it. You take every opportunity. And that’s just one of the many things that when I see that, I’m like, wow, it’s just like a startup where you know, oh, I don’t want to wake up at 3 a.m. because, you know, working, you know, with our Austrian end users, you know, might be difficult at that time. Like, no, you embrace every single moment in startup life. You just if there’s an opportunity, you go for it. And every day you’re constantly working to improve, to talk to customers, to improve yourself, improve the product, to constantly just be a better version of yourself. And that’s really what I was trying to get across there with with the fighters is like, it’s a fighting spirit. They’re they’re constantly doing the same thing, but the company is themselves, like they are their own entrepreneurs. They are building themselves and they’re no one is believing in them in the beginning. They’re constantly trying to always improve, and it just takes incredible amount of faith in yourself and in the industry that you know you’re working with. And just to believe that you know what, you can be a champion. You can make it. And that’s those are the parallels that I saw.

Harry Glorikian: Well, it’s interesting because I’ve heard Marc Andreessen describe it. As, you know, when you’re an entrepreneur, you almost have to enjoy getting up and getting punched in the face every day to the point where you actually like the taste of your own blood, and you get up and you do it again. And it’s sort of true to a certain degree. You’ve got to enjoy being bludgeoned for a while until you can make it, and even then, you know, somebody will try to bludgeon you even after that. But it does take a certain mindset to want to get up and do it over and over and over every day. Um, so look, it seems like the company is on a great track. I know that you guys are-and you’ll forgive me because I haven’t, like, read the latest—pre-seed, right? Still?

Taylor Chartier: We are considered when it comes to the definitions of stages of companies. Yes, we are considered pre-seed, but we do have we’ve had some really tremendous traction with some large sharks. We actually in our pipeline, we were approached by one VC who said that they have never seen a pipeline like we have in a pre-seed company. It’s just the way that it’s working out. But as you know, when it comes to, you know, graduating to the different levels of what investors consider, you know, the stages of the companies, they’re looking for repeatability. And as you as you know, I mean, as an investor yourself, you have all that absolutely nailed down. So we we are going through those stages, but we’re just ours is a little bit different because we have so many big partners that we’re working with in parallel. So once those all close, we’re really excited to be, yes, who knows, it might even be series A, we might not even do a seed at this point.

Harry Glorikian: It’s all about valuation though. Yes, it’s all about valuation. So listen, it was great having you on the show. I’m glad we were able to get you on and talk about this. I think it’s a it’s a critically important area. And, you know, this whole area of applying AI to images and, you know, seeing things that a human being can’t see and speeding up the process and saving money. And of course, you know, then producing a great product as a result of all this process is, is, you know, critical to, you know, like I said, making good product and producing something viable for people to buy. So I can only wish you the most incredible success.

Taylor Chartier: Oh, well, thank you so much, Harry. It’s just been an absolute delight speaking with you. And and thank you so much. This, this conversation and conversations like this really helps spread awareness of AI technologies across spaces that really need it the most, to help patients, to help people until health becomes an inalienable right. So thank you for helping, you know, elevate the conversation. And, and hopefully we’ll we’ll have some more dialogues that will sprout from this initial conversation.

Harry Glorikian: Excellent. Great having you on the show.

Taylor Chartier: All right. Thank you. Harry. Take care.