Trident Bioscience builds tools to expedite the discovery and optimization of useful proteins. By combining technologies, they're closing the design-build-test loop of protein optimization and cutting the total cycle time to help bring game changing proteins to market faster than ever before.
Loretta TIOIELA: Welcome! This is a new episode of our TechBio Founders Series by Next Sequence, and today we're super excited to welcome Tyler Shimko, CEO of Trident Bioscience. Hi, Tyler!
Tyler SHIMKO: Hi, Loretta, thank you for having me on!
Loretta TIOIELA: Tell us a little bit more about Trident Bioscience. Can you tell us how it started ? What was actually the main driver that pushed you to found the company and then lead us to where you are right now. And what are the most important subjects that you're working on?
Tyler SHIMKO: Sure, Trident Bioscience is building a platform for machine learning operations in the pharmaceutical and biotech space that we're calling mariner. And what mariner seeks to do is make it easy for scientists and ML engineers to work together to create machine learning models and then deploy those internally at a company. So, taking any sort of ongoing process that you're being collecting data on for a number of years building a model, and then allowing a scientist to share and work on it really, easily and simply.
Going back up, my background is more on the bioscience side originally, and then transitioning into computational biology and machine learning, I did my undergrad degree in biology at the University of Utah, and exited that program in 2015, came to Stanford, and thought about what I was going to be doing was very similar to a combination of wet lab and dry lab biology but kind of over time shifted more and more toward the machine learning sphere. I worked on a number of different projects from modeling transcription factor binding, using ML models to algorithms for DNA sequence design optimization. So had kind of a wide breath of experience since back in 2020, and so decided about midway through my Phd that I was working on a number of really interesting problems, and I thought there were a number of good applications specifically in the protein engineering space.
When I started, my initial pitch was that we were essentially going to help companies that we're trying to engineer novel proteins. Do that better, faster, essentially speed up that process for them. Through the course of my last year of my Phd. I had the good fortune of interviewing for and being accepted into Y Combinator, did that immediately after grad school, which was very early in my company's foundation, and through that acquired our first couple of customers for our protein engineering initial product that we were pitching.
Loretta TIOIELA: Yeah, Well, yes, that's a lot! So i'm going to start from the very beginning, because we have a huge audience out there specifically of Phd. Right now in biology, biomechanics or bioengineering, all asking question and all doing basically what you did. So I think for them it would be very interesting to unpack that bit of your story basically while you were in the middle of your Phd. You decided actually to start learning, and getting more at ease with computing and specifically with machine learning to support your Phd. I would like to understand what was actually the problem that you were working on at that point. And what was guiding you towards trying to think about bringing on ML. Into your research?
Tyler SHIMKO: Sure. Yeah. So the project we were initially working on was a novel assay to look at how transcription factors interact with different small DNA sequences. And this is something that's been done for a really long time. But our kind of new twist on it was using high throughput sequencing and a microfluidic device that we had in the lab to do this with really high resolution and getting out thermodynamically meaningful measurements of that change in binding energy which at the time was pretty novel. Now there' been a number of techniques that have been developed to do that and to approach that problem. But the problem was, we were getting out data that were kind of just okay. And it seemed like there was a way to take this massive corpus of data that we were collecting and kind of clean it up and gain novel insights from it, and some of the ways that we can do that have been done in the past. You can use linear models to address that problem and say, okay for each nucleid acid at each position in the sequence, what is its contribution to binding energy?But this was at the time 2015, when these models were becoming very popular in the biosciences, and so I kind of took it upon myself to start learning the ML side of things specifically, the deep learning side with the target to clean up this data set that we were collecting, that we knew had a good foundation, but seemed to be quite noisy and so we did that in combination with comparisons to the original measurements, to see how much we could gain from a deep learning approach versus the kind of standard more linear model-based approaches.
Loretta TIOIELA: What were you after when you say kind of cleaning up this data? What was actually not working out for you. And what were you hoping to gain with this deep learning model?
Tyler SHIMKO: Yeah. So we could see that the general trends lined up with what we expected for this well character as transcription factor, but for any given sequence we found that the noise, the variation from experiment to experiment was pretty high, so generally these large trends were holding.but it was kind of the more fine detail that we were missing or wasn't coming through in that experiment. And so we were able to use that deep learning model to essentially smooth out that data and get better estimates for each individual sequence. And then the cool thing was because this was attached to a whole wet lab operation that could measure these things in different ways. We are able to go back and actually validate that the deploying predictions were more accurate than our initial experiment.So, by combining the experimental platform with deep learning, we ended up getting more accurate data out than just the platform by itself.
Loretta TIOIELA: I totally get it. So that was when you you said you started like 2015, and then spend a couple of years to the turning point in 2020 - definitely a turning point for all of us on on earth - and you decide to go and create Trident Biosciences And how did you move from the academic ecosystem to the more entrepreneurial ecosystem?
Tyler SHIMKO: Yeah, that was definitely a difficult transition, and not only because it was 2,020, but because I didn't have any experience at the time in business, and so trying to figure out which parts of the work that I had done during my Phd, which of those scripts that I had written was going to be applicable to any new customer, was by far the most difficult part of that process, and it was just a lot of going out talking to people. We are fortunate enough again to be in YC. And so you have a kind of pre-built level of trust with a certain customer base, especially now that YC's been funding more bio companies so we could go out and had an easy in with a number of companies that were working on protein engineering projects, so that first, probably 6 months was just a process of going out, finding potential customers, seeing what their problems were, and seeing if there was any overlap between what I had already done and the problems that they were facing.
Loretta TIOIELA: Yeah, I'm going to take you on that because I totally get the idea of moving from already existing script that you have as part of the project, and then trying to map that to something that can be built into your product. So that brings me to the question of how we you thinking about who are your customer, and why they would be interested the way they understand it. When you explain it. It's a little bit like : “Okay, I could see that I could do that better with machine learning and let's try to turn this into a product” That means that for me, if i'm just basing my observation on what you just told me, your primary customer would have been research scientists, you know, people that potentially, were trying to do exactly the same thing that you were doing in the lab, but that would not go through the hustle like you did of having to learn machine learning just to achieve this objective? Was it the way that you guys actually started? And did you stay on that as a primary target? Or did things like move along a lot along the way?
Tyler SHIMKO: Yeah, there's been a lot of movement and pivoting. As so we stayed on that initial course for about another 6 months following what I see working with those initial customers, but I should say those initial customers were all very small biotech startups, as you would expect, coming out of YC. Which I thought was going to be our initial customer base, kind of, because they're a little bit easier to sell to. They tend to be in groups that are a little bit more research focused, and that was definitely the case. The problem that we encountered was that for each of those projects it becomes a very deep niche almost immediately. So you start working on the project and you get sucked into the new tools that the company wants for their specific application, their specific target. They want to model what works for their protein family. They don't really care if it works, for, you know, if they're working on an enzyme, they don't care about other enzymes. They don't care about antibodies.So for us to build something that worked both generally and in every use, case was almost impossible.So we had the good fortune, as we worked on those initial projects, to apply some of the graph neural network models that we were building actually to small molecules for a large pharmaceutical company, so that represented a pretty big pivot in what we were doing, but also the speed with which we could move. The large pharmaceutical company had a lot of the data available already, and so we didn't have to wait to go back and forth between the lab data collection, and then we retrain the model and then go back to the lab. It was much more of a kind of standard machine learning workflow of collect the data, clean it up, analyze it, run the machine learning model and then back test against the most recent data points. And then we could give them a pretty quick analysis of how well an ML model is going to work for their application. So, starting about the end of 2020,we really started to focus more on applying the graphical networks that we were working on to those small molecule problems for larger form of customers.
Loretta TIOIELA: So yeah, you definitely, actually change entirely the way that you were going to position the product to move away from working with small startups that basically had the same DNA, as I tried, to moving to larger paying customers that actually add the data. That's where they required it, any way, to be able for each model to really move at a more general model level able to address these big use cases. And so you did that work with the new type of customer. Specifically, pharmaceutical company, because traditionally phone mark can be something of a kind of a based, for since it's a make or break situation. And same as I did, you know, you start integrating with your team and these team that have totally different kind of mechanics and processes. How did this integration go?
Tyler SHIMKO: Yeah, there definitely been some high points and low points along that way. I think the biggest struggle for us has been the speed, so we can obviously move a lot faster. This, for us, represents especially initially our only client, and even to this day one of those small number of clients, and so we could focus really deeply and and make really good progress quickly working with that customer, whereas, if we needed anything from their side. It was often kind of months of waiting, trying to find the right person, making sure that you're in contact with that person at the right time for them to address your problem. So that's been the biggest pain point. But I would say the bigger benefit is the ability to move internally within that company, and that's actually how we arrived at our current product. The machine learning operations platform was recognizing that we couldn't just build them a model and ship it to them as a python script. A number of the scientists don't have experience working at the command line interface, and they're the ones that are most going to benefit from that model. So, through a number of presentations, feedback sessions, conversations that we had not only with the group we were working directly with, but also people above them, people moving laterally within the company to them, we recognize that there was this opportunity to just take machine learning models and make them usable for your standard everyday scientists. And if you could build models based on internal data even better, so something that you could take an off the shelf model, attach a user interface, make it really easy and simple for that person to use, and then they could work that into their daily routine for any given experiment, and that was where the real power was going to be in the real value for our our large circle of customers. So it's been great because we've learned a ton about how that R. And D process works where ML fits in and where the pain points are for the customer. But yeah, there's definitely been some a bit of a learning curve in terms of how quickly we can work. who the right person to talk to is, how to manage billing and finances kind of all of that was new, working with a larger form of customer. It wasn't all direct one on one conversations with the CEO or CSO like it was in our previous life.
Loretta TIOIELA: Yeah, definitely. I can only imagine one of the things that you said which really like intrigue me a lot is the data part, because obviously you need access to these data to be able to train. And so how did that go? And what was the process that you put in place were you granted access to these data? I can only imagine that these data were not moving around. There's so many rules out there, especifically, I would say privacy concern, legal concern. And so how did you do?
Tyler SHIMKO: We adjust our way to that to be able actually to, within these constrainst, still be able to move away from the scripting part to actually build a product.Yeah, so thankfully, we were able to avoid working with any protected data. And so the data was a much lower security level in general, in terms of kind of jumping through their security hoops. There were a number of written reports we had to provide. We had to undergo a penetration test for the infrastructure we were going to use on their data, and we they wanted their data pretty well siloed within our cloud compute environment. So we were able to go through that process with them, and it ended up being much more congenial than I expected initially. So once we got in contact with the right person in their IT department, we were able to move pretty quickly and check which of those boxes were relevant to us, which ones weren't relevant, and just make that process kind of as painless as possible as it could possibly be. So we were very fortunate that we were working with kind of lower security data, and we could move it onto our infrastructure after we had met all their security requirements. I think my recommendation to someone out there trying it is to try to find that right person because that was the difference maker for us.
Loretta TIOIELA: And what kind of a team did you have in front of you guys? Because they can only imagine that moving away from small startups that are basically organized in the same way that you are originally to a big pharmaceutical company, it's definitely a game changer. So were you dealing with the team in charge of real data evidence, where your day to day, you know, working with peers on the pharmaceutical company side?
Tyler SHIMKO: Yeah. So it was a a research group working in one of their process chemistry departments. So it was mostly scientists and not a ton of ML engineering experience on that side, running your standard assay types for that type of small molecule chemistry.
Loretta TIOIELA: And so you were, basically, it's very interesting, because you're basically an interacting directly with your end-user which is different from your customers.And so these end users were basically researchers, which was very interesting because you could target, I would say, a element that you had to build into the product to make it work for them within their workflow. Like you mentioned the need to have one interface, which is very simple, engineer friendly. On your side your structure was able to support these chemists, you, yourself being, I would say, researcher. I think it's actually a very huge value out, because you can speak the same language. But what about your team?
Tyler SHIMKO: Yeah. So for for the first part of this project, it was just me. And then I realized pretty quickly, especially when we moved more toward the user interface side that I was out of my depth there. I've done a bit of web development in my past, but not nearly enough to provide a a professional product. So fortunately I was able to find contractors that were able to fit that bill. They've been incredibly talented and contributed, not just in terms of the code base, but also in terms of the intellectual side. They've noticed and made changes to the UI that I never would have recognized as important or valuable. So I've been fortunate to work with a great team of contractors to make that side of the product work, and so the split tends to be a little bit more me on the ML and research side, them, mostly doing the kind of app development and web interface design, but slowly moving more and more toward ML. Some of these routines we're starting to kind of codify and make a repeatable process out of. So it's been great to see them start to excel on the ML side, as they understand, you know, kind of the problem design, and how we want to frame any given new project that we take on from other pharma partners, moving on to the end-user i'm actually interested because the core product itself is a software product, right? It's a ML ops software dedicated to chemistry.
Loretta TIOIELA: You work on designing that ML product to actually fit these chemist needs, because you are in chemistry you have so many element of the workflow. You cannot really actually think about realistically build one size fit all platform and all of the workflow. So I guess one of the part and the decision process important in that product development must have been to really identify which part of the work for were you going to be the one you were focusing on.So how did you decide that you were going to do protein design engineering. And how did you then move on to? Okay? What is out there on the market, and why my product is actually radically different. And what is the value of them bringing on there?
Tyler SHIMKO: Yeah. So the initial decision to go into protein design was primarily based on my experience in grad school. So that was kind of the only thing I knew the thing I knew how to model relatively well. And so we decided, You know that's gonna that's what we're gonna start. The move into more of the chemistry focus side was just driven by essentially market demand. Our customers wanted that and they were willing to pay. And so we knew we could apply the same types of tools to that different problem set. And along the way we could learn how to potentially add to our tool. So we, you know, started off in protein move to chemistry. But then that process of identifying what the problems were in chemistry that we could actually address at that was entirely through this back and forth process with our partner.So we were able to kind of get off the ground, get them a model that worked relatively well, initially. And then the process became a back and forth. How do you want this presented? What type of UI do you want? Do you need? Are you trying to do this at a certain scale where you're trying to make many predictions, and we need to batch them. Are you trying to just do one off predictions?
TylerSHIMKO: And we realize pretty quickly through that if we broke this down we could make it kind of modular system for just the the ML model side of things. So we don't focus on modeling every step in a chemistry process we focus on just you have a process in place, and you just want to take the inputs and predict what those outputs are. So by defining the limits of our platform we are able to focus just on which parts of that are repeatable. So for things like molecular, input we know, we can take small strings we can take of molecules and input those into a model and return a prediction. So by breaking it down into okay, what are the things that all of these people say they need? Can we build those in a repeatable way, and can we deploy them all as one unified product without having too much human interaction. That's how we kind of got to this place that we're at right now.
Loretta TIOIELA: What was the original you know inside from them that made them look around because these pharmaceutical companies, they have team specific, i would say overspecialized in different area. And what was the input for them? What was the interest for them to try to out these, to another startup, try to find another product that could potentially be on the shelf like yours, but that they could use.
Tyler SHIMKO: Yeah. So I think the the main incentive for them at the very beginning of this project was just a a cost and labor savings. So the reason they wanted to model for this process was to rank the best way to analyze a model or a molecule. Rather so they wanted to be able to save the effort of having a technician run that assay repeatedly, and they wanted to save the cost of the materials for unnecessarily running experiments that you potentially knew weren't going to work based on the data that you already had. And they were sitting on this kind of dormant data set that had been collected over a number of years. So they realized there's probably some potential there to make use of that and save themselves work and money in the future. So that was the initial incentive. And then I think, kind of where our partnership really benefited from working with them was realizing Okay, what is the end goal here for this model for this project? And how do we need to present that to the end-user to make it the most valuable.
Loretta TIOIELA: And you have an idea of what kind of market size it represent, because from memory, at the time when it was working on the problematics of data, pipeline and data modeling for small scale companies. I could see that they have indeed a huge problem, which is, there are indeed costs. Basically we present 20% of their revenues. So every year on, they actually invest 20% of what they're getting into the R&D pipeline, and that includes everything. So what part of that? R&Dpipeline actually represent the pure chemistry part? And what did it meant for you and for your business in term of business potential.
Tyler SHIMKO: Yeah. So I think it's an interesting question, and one that's been hard for us to address ourselves just because it's growing so rapidly. And I think we're also going to see kind of a solidifying of which areas of chemistry or biology, ML is most applicable to, and which divisions within your company are going to benefit the most from that application. So I think it's hard for us to say kind of what fraction of that kind of pure chemistry side can benefit from our platform right now. But what does seem apparent is that for a lot of these, especially large companies, they're building out ML teams internally, but those ML teams are understaffed to be able to meet the demand for all of this latent data that's sitting around. So if we can position ourselves and make a product that makes it easier for a team of scientists to within 10 or 20 min of an ML engineers time build a model that that works really well, I think we'll see an explosion in the demand for those types of models. I think the demand is there. It's just kind of unmet by these these internal teams currently.
Loretta TIOIELA: Yeah, I love this because that was basically one of the thing that really seduced me originally. When I got a glimpse of what you were doing, and it was like. Hmm. This is very interesting. I need to follow this because it kind of fits the other whole idea that I have about the move from biotech to tech value. Now, in techbio there is totally different approach. I would say conceptually, because where before I've seen specifically in protein design and emulation. You have all this logic where you're going to find a service provider that is going to attach itself to one specific use case again, as you said, it's very difficult to be something that is going to work for any use case in life science, the complexity of biology, in general make it almost impossible.But what people over the years I've see doing is they build something that is very vertical, meaning it's very aligned for that use case. But it's not. I would say software per se, because even though they might be using software it is not with the logic of something which is going to be basically a one size fits all tool that can be used by different team, and that basically enough to show product, whereas in techbio we can really clearly see that founders are not making this mistake anymore, and they understand that. But they're building is actually a piece of software that is going to be used by this team internally. And so that's actually for me represent the shift that you were mentioning the pharmaceutical companies, now understanding the radical shift that it's happening, the building slowly their own internal data team data factory, ML team to be able to do that. And so, instead of you, providing I would say a piece of software that we do everything for them for that specific use case. You totally understand that they need not a vertically integrated software. They need that one piece of software that would be the tool that their ML team. When the ML team would be ready, would be using again and again. And that's very interesting, because that mean that you can do that for this actually specific customer. But tomorrow you can extend that to any other customers, and if you really want it to, you could do that for typically any kind of obviously customer that fits this model, and that mean that basically you could open source it. And you could become the communities reference of protein design or specifically, the ML Ops for protein design in chemistry, in that specific angle, so that for me it's very radical, radical, shifting and understanding, and the way that you position in your product.
So how did you manage to talk to investor? Talk to people that are going to come along and support you because we're not talking about the classic biotech product here. We're talking about the tech bio product. And obviously there is this kind of software approach, and we're talking about ML Ops.So what were your way to introduce a ML Ops to people that didn't know anything about?
Tyler SHIMKO: Yeah. So I think it's been. It's been very interesting because we've kind of been surviving on the initial raise. That was a very different product. But we are gearing up to talk to investors shortly in the future. And so this is a problem I'm definitely going to have to address then.So I think that the biggest way we can do that is, by taking the approach that I would as a scientist of you know, we are building a pipeline from these flexible components, and what we want to do is have one that works for us at the very end.So I think the main way that we can address that is, by kind of telling our investors that you know there, isn't just one way to do any sort of analysis in the bio sciences. There's many different ways, and you often need to mix some at the puzzle pieces that you're using in order to get to that destination.So for me as a scientist it's an easy concept to understand, because you know, as a grad student I was constantly taking tools, linking them together, writing glue, code. And so the pitch is easier for the pharmaceutical side than for the investor side. I think the challenge is always going to be you know how big is this market, really, and if that market size, you know, is going to be divided up then among Trident and every other one of our competitors or complementary tools.
How do we make that pitch? And we're still very much figuring out how to do that. But I think the massive growth of that segment is going to be helpful moving forward, just mentioning. You know, that this seems to be taking off it doesn't seem to be slowing down. There were people that doubted it in 2015, and said that it was just a fad then, and they're definitely overblown parts of it. But on the whole, more and more companies are adopting it. More and more companies are using ML for their applications.
Loretta TIOIELA: Yeah, definitely, I would say that when I started 5 years ago, definitely, people were basically thinking that I was crazy I think. Post 2022, definitely. People are not saying that anymore. They have been in 2022, some very defining moment. I would say, of course, like for any industry it's true now but I think we've seen it again and again when you have market players making significant shift.
This is the way to signal to the entire ecosystem, including more general as people, or, I would say, the general public in general that something is happening, and in 2022, I think the work tremendous work done by DeepMind has brought, you know, aground breaking, I would say, awakening about what was actually happening for a lot of us. We have been working on this for many years. It was no news, but it's felt good to be comforted that now something is happening. The timing is now, and it's not just a fad.
Most of the time investors ask the why now? what's different now? And I think that that kind of signal are actually very interesting because they allow startups founders to really be able to, you know to aggregate behind that to be able to corroborate what they knew deep down inside was already happening, and so in the ML Ops. So i'm an obsessed, has been kind of this kind of this world that it's been going crazy over the last couple of years. How did you position yourself? And if you had like a lot of other team working on the same idea of bringing an obstacle chemistry, or do you see it as something where it is still uncover right now, and there's a lot of fun for you guys to actually play around.
Tyler SHIMKO: Yeah. So I think it's still you have large open source projects that are moving into the ML Ops space and definitely serve as a kind of one size fits all solution, and the options for these teams at the companies tend to be do we use one of those off the shelf, either open source or commercial solutions that may not be directly a fit for what we need to do or do we either build something ourselves or find a more specialized product. And for us, as a small team as a small startup, we're never going to be able to directly compete with some of these larger scale, like ML flow projects. But where we can fit-in is in specializing those projects, or making our own projects that are directly adapted and built for either biology, chemistry, the unique challenges that exist there, because, no matter how big those open source or commercial projects get the larger market, and the easier demand to meet is always going to be in things like computer vision and natural language processing, and so I think that will serve as kind of a large enough distraction to minimize the competition, because there's also a high barrier to entry. I think it's difficult to do this well, without at least some sort of background in biology and chemistry to understand what those challenges are going to be, what things are unique to am I operations specifically for pharmaceutical and biotech companies, and how to execute on that. Well, there are tons of projects that have met the dustbin, because, even though they had a great UI or a great user experience, it just didn't do the right thing, and so to kind of piggyback on your mentioning the work by DeepMind and others to build these massive models that have been very successful at predicting, I think one thing that especially investors need to understand is that the number of projects where all you need is Alpha fold is very small.Often you're trying to take a prediction from Alpha Fold and combine it with some of your internal datasets, tweak it, maybe fine-tune alpha fold for your specific protein family, and then use some combination of those tools. And that's where it gets really difficult. And you end up having to bring in a number of ML engineers to do that, to vet the data, to build that model, to tweak it, to train it on GPUs, and then deploy it to a customer. And if we can automate that headache, even with smaller, simpler models that gets us a lot of the way towards solving the real problems that we've seen in at least R&D
Loretta TIOIELA: Yeah, definitely, yeah. Totally agree with that. What about the future? What are you guys thinking about? How do you think about the future. What are the next steps for, you guys?
Tyler SHIMKO: Yeah. So right now, we're trying to kind of finalize our initial feature set for the platform and get our initial customers on board. So we've done a lot of contract model development work, and now it's pushing toward okay, which parts of this are repeatable. What do we need to get into it to make it functional and useful to an end user and hopefully bringing on in the next few months, our first one or 2 customers, to really start kind of testing it out and informing our design process. Moving forward, we have a lot of exciting ideas about how to either take off the shelf models and fine tune them on local data sets, build it models the novo do some sort of kind of suggestion procedure for how to build the best model for a specific application.But all that kind of depends on our customer demand. And right now the most important part for our customers is just getting models into the hands of their scientists.
Loretta TIOIELA: Amazing, amazing! I'm definitely looking forward for many reason, I think, in a certain way, the way that you have approach thing is very similar to what I was coming from at the time I was working on Alzheimer disease, and we were looking at our blood biomarkers, and I had to go through that same process of trying to identify, what can I do to build better model? What can I do to have better access to a better data. What do I need to set up in place to have this highly optimized and so i'm definitely cheering for you guys. And i'm looking forward to what you're going to be building specifically, because there is something that is happening that I think needs to happen in biotech, which is basically the fundamental shift that happened for computer science 30 years ago. 30 years ago we had proprietary software, but it is opensource that brought us so far. But at some point we needed to move along and then open source community software development happened, and I would say that everything that we have right now is kind of the product of that. And so i'm very looking forward to seeing that happen in life science, and for that we need specifically 2 things:
We need to see more product that I understand that they need to focus on one very key area developing actually tools.
And that's where building actually software ML Ops platform, specifically dedicated to specific use cases in that of the whole entirely over complex element that can be life science is actually critical.
Life science is moving, and towards what? Up and up, and since a couple of years now towards computing. And so it's very interesting to see the dawn of it, with companies such as yours, and so super exciting and cheering for you guys. Any last words on how best to contact you. If people are interested in the product. Or if people are interested in talking to you about the future of the business and what you're planning to do next.
Tyler SHIMKO: Sure the best way to find out more about us is to go to our website. Trident Biosciences, or you can email me directly at t.shimco@tridentbiosciences.com, and i'd be happy to talk to you. Any potential customers interested, users with feedback thoughts, ideas, kind of open to any feedback at this point
Loretta TIOIELA: Awesome! Thank you a lot Tyler, and looking forward to see what you guys are going to do next.
Tyler SHIMKO: Thank you so much !
Drugs & Therapeutics
This is where we deep-dive on the most exciting trends in #drugs&therapetics
This is where we deep-dive on the most exciting trends in #drugs&therapetics
