Avantor, Inc. (AVTR) Earnings Call Transcript & Summary

Hello, everyone. Welcome to BioProcess International Spring Digital Week brought to you by the producers of the face-to-face BioProcess International Events. My name is Barry Walsh. I'll be your host for today's session titled, Digitalizing Biopharma: how long material data can maximize biopharma supply chain in manufacturing practices. First, I'll cover some quick housekeeping items. If you experience any difficulties with audio advancing slides, refresh your screen with F5. If you are still experiencing other issues, hit the question mark button to receive assistance. [Operator Instructions] Now let's begin by introducing our speaker from Avantor, Claudia Berrón, who is the Senior Vice President of Business Development and Commercial Operations of biopharma production at Avantor. Thank you for joining us today, Claudia. Now I will hand it over to you to begin the presentation.

Thank you very much, Barry. Good morning, good afternoon, and good evening to everyone. Thank you for your patience as the event team worked here quickly to solve some technical issues, so happy to be live. I appreciate each one of you taking the time today to talk a little bit further about raw material data, how we use it today in the biopharma 4.0 journey and how it can help influence the industry today already and in the near future. As Barry mentioned, my name is Claudia Berrón. I work for Avantor in the commercial operations and business development for biopharma production business unit and happy to be talking to you about today. So what is it that we're going to be covering today? So the key thing, what are we going to talk about? In the last, let's say, a little over 5 years, right, biopharma has come a very long way when it comes to production data. We are leveraging what's now statistically relevant data, captured via a lot of online monitoring tools. And we're starting to do data learning and e-learning for process improvement. There are now a lot of very solid software platforms that offer the end-to-end workflow. And most of these tools provide very good analytical add-ons that -- which include raw material capturing and then linking of these raw materials to the drug product batches. But now let's focus on how do we assess raw material variability, where all of these tools are now enabling us to link raw materials back to drug product batches, the focus on raw material variability continues, right? How do we, as an industry, focus our collective efforts on improving the raw material data mining across the supply chain? So as we know, talking through the slide here to through the left, in biopharma, we know very well did the product is the process, right? The process is a very critical part, and that's what defines the product. We have an incredibly complex supply chain of raw materials, which is something that we have to think through as we think of raw material data mining. Our manufacturing is -- lead times are long and costly. And then it's very difficult to change, right? So we have to assess once the process is developed and ideally from the beginning, how do we build-in the raw material availability to meet our processes. So as we think through this, this is what we're going to cover today, right? If we can increase supply chain visibility, leading to shorter investigations, most definitely start with reducing, but hopefully anticipate variability that can lead us collectively to a lot less investigations. And ultimately, right, if we very well understand variability because we have to come to terms that variability is not going to go away. But the better we understand and, ultimately, as we develop and manufacture drug substance batches, we can start thinking about elimination of [batch] failures. So we're going to cover 3 areas today, and I'm going to walk through some of these on a quicker note to leave time for questions at the end. We're going to cover process developments, right? And what is the role of raw material data in it, supply chain, very critical to our industry and quality. So as we think about -- I'll have just one slide on a Avantor on who we are and what is it that we do, and then we'll jump into the content. So before we start there, let me just walk you through. Avantor, we've been supporting customers from discovery to delivery, meaning from the last stages all the way through drug substance, production and drug product excipients for many years. This slide provides a little bit of a coverage of what it is that we do. Throughout the full process, we supply customized single-use assemblies and on-site services as well as supply chain services. Our J.T.Baker brand is well recognized and specified in over 80% of the top 20 selling biologics today. We have -- we focus a lot on upstream, very high-purity and well-characterized cell culture supplements. And the J.T.Baker brand is well-known also in downstream regulated process chemicals and excipients. We work very closely, as I mentioned, with all top biopharma as well as with many biotechs through multiple alliances globally. And we're in a position globally in all 3 continents with GMP manufacturing facilities. So with that, let's move to the first area that we're going to focus on, raw material variability and the role this plays in process development. So this is a standard process development map. Most of you would be very familiar with this as we define a process. For drug substance manufacturing, there's quite a bit of elements that take place. The manufacturing consistency in biologics definitely hinges very heavily on the process itself, as we discussed earlier, but also on the raw material variations going into this process. So identifying these raw material variations really early in the process is key to define, and I know many of the people on the phone are focused on doing this over many years. But think about the role that now data is playing and artificial intelligence, AI as we develop processes, right? When we think about raw material variations, we're always thinking of the high- and low-end parameters, right? And using AI and predictive modeling to help us design the best process possible. Can this process handle the variations, right? Will we need to adjust the process to accommodate for some normal variations in this raw material? Or do we need to start thinking, is our process going to require some custom specs, some things that we are going to have to control for that particular raw material to make a process consistent? While this has been going on for many years now, there's always the challenge of how do we get statistically relevant raw material data to feed into all of our models, right, to enhance our predictive process design of predictive analytics. And on the flip side, how do we link all of these inputs and all of these heavy statistical raw material variation data to the output -- to the process output? So as we think about at Avantor, we started thinking around this challenge. We started thinking around, okay, where can we, as a supplier, help the customers assess where do we fit into the process and how do we help them define. So we started looking at different molecules. And then where will we position? Each of the little blue dots there would give you a little bit of as we think about where we position in the bill of materials. So as we think of that for every one raw material that a manufacturer and a biopharma manufacturer receives or the couple they purchase every year, we, as a manufacturer, likely manufacture 30x to 50x the number of lots, which means 50x the number of relevant statistical data that we could be -- make accessible to our customers in these efforts of process development. As we talked through this, they said, okay, complex enough. So what other raw material data sets are available there? And what is -- might be of interest to the industry? So this is a very pictorial way of depicting as we started thinking about, okay, this is, okay, a natural drug. How many times do we touch it? And as you start thinking of raw material lineage and what is the raw material data sets behind it, we started looking through the multiple feeds that could enter and feed into an AI learning process. Today, we do -- raw materials you have actuals, right? And you can see the multiple ways this drug is touched. There's a couple of cell culture supplements. There's buffers, there's a chroma resin in there. There's multiple excipients, and there's a couple of single-use assemblies, right? For Avantor, we're focusing right now on things we manufacture, as you guys know, we distribute a lot of things, which is the left side of the screen. But we started to focus on the manufactured products. The reason for this is because we have a lot more relevant data that can enable AI learning. So if you think about where are we as an industry today, likely some of you are, likely some of you are not, using and sharing already transferring e-raw material actuals, right? This is where we are today. But as we start thinking of what's behind it? For every regulated material or chemical -- process chemical, we run stability for GMP materials. So there's a lot of interval -- stability interval data. For all the manufactured products, we have quite a bit of manufacturing. We do a lot of in-process testing during manufacturing, which provides us a lot of data. And very interesting and where we have found most of interest from our customers is on the CFA of what we call the raw material components or the N minus one. So our raw materials were -- that feed into materials that we eventually supply to our customers. So there are multiple levels of variability accessible today, right? And the key and the challenge is, how do we access this in real time? How do we help this -- enable this to help us predict process development? And then how do we link all of these raw material variability into particular drug product batches, right, and into process performance? What we have found and we've been working with customers closely is some of these have been started to link to specific technologies, let's say, can we start linking some supplements up to your maps, right? Or can we start linking some of these to recombinant vaccines? And enabling that and then going all the way to the specific molecule. So where do you start, right? I've shown you kind of an ocean of data. So the question is would this all -- this ocean data set on the previous slide, where do we go from here, right? One way to start is assess what your area of focus is, right? What are the levels of variability and what are the impacts of that variability it's in your process? So for -- we took a stop, for example, for the product categories resupplying into the industry. But are these levels of variability, red being high being very variable. And what is the potential impact it can have in the process? So as we think around cell culture supplements in serums, very high level of variability, right? Significant impact, right? There's inherent variability in some of these -- a lot of these are biological sources and impact to variability is very significant in upstream yields. As we think about the buffers, right, and as I've talked to this with many customers, even though it's yellow, we have seen buffers have a significant impact in particular processes, right? So traditionally, well characterized, traditionally in the medium, but it might vary by process. Chromatography reference, it's typically very well characterized, as you guys know. The impact of this can be catastrophic downstream in terms of purification. So there is a significant potential impact here. Excipients, even though it's yellow, it's because we, as an industry, have spent laser-focused attention on characterizing excipients and understanding variability of excipients, right? Doesn't mean that there might be a variability in issue. And as we know, getting so close to the patient and the risk that it involves, the impact that the -- variability is very significant. Single use -- we see, and then you guys we're happy to discuss this. We see a level of variability low, right, and impact of variability equally low. This has been through many as single-use has developed in our industry. They're well-controlled and characterized as we move along. So how do you start, right? So some of the things we've been discussing, right, with the industry is, do you want to start with your high-touch products, right, products that touch many of your molecules? Or do you want to do with your high volume products, let's say, you have a volume that you buy significant volume stuff. Is that where you start, right? Do you want to focus on a particular process step, is upstream the focus, right? Typically, we see customers starting with excipients in film finish as -- and moving up in the process, right? But it depends on each of the companies, where do they want to start? And the key component here across all your supply network, right? E-data sharing and statistical relevant data, enabling the supply chain to transfer this, is going to require a significant amount of engagement with your suppliers and collaboration for all of us to get there. And I'll talk a little bit about there, what we're doing in the industry. So let's move to supply chain. Supply chain, as we know, incredibly complex supply chain in our industry. I'll walk very quickly through this slide. I mean through the supply chain, boy, do we have amounts of opportunities to share e-data, right? There is a very active collaborations going on, on delivering supply chain data or e-data, right, and opportunities to transfer, to utilize the supply chain for raw material data transfers. And I'll talk a little bit about more on this side on the next slide. But what is it that we're doing, right? So think about your supply chain, I'll walk through the current today. And through the current means, some of us are doing this today, some of us doing this to a less extent, but it's something that is available and current in the industry today. So as a customer issues a PO to any supplier out there in the industry. Today, there's quite a bit of companies that want to inclusive, they utilize ASN or advance shipment notices, right? In advanced shipment notices, this wealth of information that can be transferred electronically from -- starting from our lot numbers to CFA data, to supplier lot numbers, manufacturing lot days and then linking it back to our customers, right? So the ASTM standard for e-data that were part of a collaboration of the industry have serviced as a reference guide to get us started in this space, right? So you place your PO, you feed all the ASN data that's required, right? And we take that and put that in ASTM standards. As a manufacturer, we manufacture to the specs, label, package, et cetera. And as the product ships, right, typically, the system transfers the ASN over to -- typically through a third-party software and then over into our customer feeds, right? The ASN includes, as I mentioned, could be a wealth of information, right, including transportation carrier. For biopharma customers, we do optimize quite a bit of the field, depends on the customer that they want to see in the ASN, depends on what the system requires, depends on how they manage their inventory and depends on how far along are they linking the ASN data to their process and material needs. As this comes into our field today and to your warehouses, let's say, it gets automatically loaded, right, it's transferred into the warehouse management system. And automatically, you intake all of your CFA data, right? So you have a wealth of information from the -- particularly lots you purchased. So typically, you scan it and then you know what you receive, exactly what you receive and you see a base in your system already. Now let's start thinking about the future, right? Where could this evolve? And we can think about this futuristic or you could just think of your package, you ordered from Amazon, right, because this happens every time. So as we start evolving in this space of biopharma production and the supply chain data, we're starting to think about how can a PL be automated manufacturer, right, as you're scheduled and the runs for reactors are scheduled, we're way in advance, right? So all your bill of materials are fed into your system. So as you put your scheduling, this would -- starting to evaluate, can the system trigger? Based on the BOM you want to run, you automatically trigger your PL. ASTM standards, while this has been very, very helpful to help us get going, they're not fully defined, right? There's a lot of things to work on. There's a lot of things that are optional. So some of these are loosely defined. So there is collaboration that needs to happen, but fully defined standards would be the future looking. As you think about when you get your ASN, can you link your ASN to your manufacturing run? A lot of softwares that we utilize today trying to connect raw material lot to a particular drug substance batch, right? Can the system manage this automatically? Can you exactly know, within your system, every single raw material lot that has been used with the CFA data that goes with it. Now as we move into shipping, right? And as I mentioned, this is not rocket science. You can go into Amazon and track where your package is every day, with dynamic ETA calculations. So this is something that is done in the biopharma, we're exploring how to move in that direction. As you think of quality, right, automatic quality, intake of all your CFAs, right? And we'll talk quality as the next topic here. And as you think about -- once you have a knowledge share receipt, right, could you notify immediately, you're manufacturing. I have all of your BOMs ready to go for your run that is due to come in 3 weeks, 4 weeks, what have you, and this is your inventory stock, and this is the lots you have, and this is how the variability of the incoming raw materials that we have plenty of data analyzing, potentially links to your drug product batch. So it's a -- kind of looking at what could we do in the future, right? If you think about how we're doing today and how this evolves, the other piece that has become quite interesting is in the supply chain, right? Going into pretax, right? And could be -- the liquids could be powders, we have a powder example here. But as you think of incoming raws and enabling faster release, right? Any incoming [indiscernible] we received with an automated ASN, if you are thinking about pre weights or evaluating, don't forget about the e-data, right? This is quite important, right? So typically with your ASN, if you go and discuss with your suppliers, I want to get prepacks ready to go, lock-and-load raw materials in my prepacks with exact weights, typically, your ASN would carry your weight ticket. So not only would you save time, right, significant time in ID testing because of Roman ID capabilities and Safe Time and QA release, but you will automatically from the time you receive your ASM, meaning from the time your supplier ships your material out of the door, you would be getting into the system the exact ways in each of the packages ready to load into your process. So this -- and there's probably been quite a bit of studies around prepacks and what does this do, right? So basically, we're talking about going from 30 hours of taking your traditional drugs and weighing, dispensing, QC testing, et cetera, to maybe 9 hours of you getting your raw materials in prepack, ready to go, right, receive the samples, you do Roman ID for most of these, which on the Roman ID, we'll talk on the quality side because that's really really important for e-data. And then the incoming ASN, with your pre weights, you're automatically feeding your warehouse management system as well as your manufacturing box, right? So this is a wealth of information. As you think about the complexity of our supply chain as how do we streamline our supply chain and how do e-data helps us really have seamless supply chain between suppliers and manufacturers, there's a wealth of areas of data to think through in terms of e-data. So we'll go to the last category. We'll talk a little bit about quality. And I think I hinted on the previous slide, what is it that we're going to talk about quality. When do you think of e-data. We have been, as an industry, doing Roman ID for many years now, right? And the benefits of Roman ID are well known, right? You really reduce your incoming test from, let's say, 2 days to less than an hour, right? But what is often overlooked when we do this Roman ID testing is what is it that we're capturing, right? And how can you talk to your suppliers more around what is it that you have in e-data from your Roman libraries that can help us in the focus -- laser focus that we have today on raw material variability that is an other, very wealthy bucket of information that is oftenly not -- often not tapped in our industry today. So as we think about Roman ID, right, the biopharma manufacturers do this, suppliers do this to create our spectra library for Roman ID, we have to capture very rich, very heavy sets of e-data. And with this, the variability that comes in the raw materials, right? So typically, right, you're a supplier, and it varies by company. So typically, you would take 10 lots of each of the products to do your library. Each lot would do triplicate, right? So to assess that your library is reading correctly, from a supplier perspective, because we service a wide range of customers, you typically would do different types of liners for solids, different types of glasses for liquids, right? And you would have a wealth of data, and then you can establish your control based on your packaging materials, right, and your Roman. Again, we do this from a quality perspective today for incoming testing, right? But we often do not tap this information from a raw material variability perspective. To think through, there is a wealth and richness of information here also that is ready, first, which is very key. And then ready to be tapped in terms of raw material variability. And there is a wealth of information after suppliers. We're capturing the data. So going back to the topic of collaboration and engagement. Supplier engagement and collaboration is going to be really, really key to get us, as an industry, to the next level of e-data sharing. And what does this e-data do in our process development, in our trouble treating efforts in manufacturing and within our supply chain? So I'll summarize here a bit, and then we'll have some time for questions. So as we think of raw material e-data, how do we help transform our industry, right, productivity? How do we help on the beginning, develop process -- develop the processes much quicker, much more efficiency in understanding the raw material effects or the raw material variability effects that could potentially have into our process, right? How do we take from a supply chain perspective, right, all the richness of the raw material data and make our life and challenges much easier from a supply chain, if all of us know exactly what's in the packet, how much is in the packet, what is the way that I can receive it, trigger from a bill of material planning a PO and then know exactly what I'm getting in and be ready for my runs, right? And from a quality perspective, how do we capture all this data, right? How do we capture all this data and then make sure we utilize it, right? So there's a wealth of opportunities, right? And that's -- I'm very excited to convey this because we've been working a lot on this, and there's so much opportunity here. The data flows, right, will most definitely change the way we do business and the way we collaborate, right? Data richness in our space, and you saw that slide that had the complexity of the amount of data, right, will create new opportunities to make our industry more productive, not only in the supply chain side, but on the manufacturing perspective. Wealth of information, ocean of information, where to start, right? It's always a challenge, it's always a question we get, right? Start with -- I cannot emphasize this enough, collaboration, right? Start talking to your suppliers, right? Collaboration will be key to a successful e-data platform, right? Shared data platforms, right? How do we provide data, real-time data access, right? We've been giving a lot of thought about this at Avantor, right? As I mentioned earlier, if you as -- one of some of our customers buy one lot, but we manufacture 50, right? The richness and the statistical relevancy of that 50 lots is very important. So how do we allow real-time access to the data? How do we continue to collaborate, right? We've gone a long way with the ASTM standards. How do we continue to collaborate in full implementation of the data standards, right? And then how do we start digitalizing, right? Raw material at all levels, right? So I mentioned to you, CFA, right? That's the current status, right? There's all this data behind, right, going down to the N minus 1 raw materials that is accessible. How do we digitalize all this. And then very important, how do we as suppliers and with their customers have dedicated teams that whole -- work very closely, not only to mine the data, but to share the analytics, right? The data is only as rich as the analytics you generate from it, right? And we all know that, right? And as we start going more into the journey, we're talking a lot about the raw material data richness, right? But how do we take the data then, right, and really turn it into insights and analytics that can help us around our wheel here in process development, supply chain and quality, support our process and take us into the biopharma 4.0 golden batch. So with that, I'll stop. Barry, I'll turn it over back to you, and we can move to Q&A.

Thank you, Claudia, for an excellent presentation, and thank you again for your patience while we were experiencing technical difficulties at the beginning of this call. To make up for that, we're going to extend the question-and-answer period by several minutes to make sure that we can accommodate all the questions that will be asked of you over the next 15 or 20 minutes. So we do have a few questions that have been submitted already, but let's give the rest of our audience a moment or 2 to enter their questions in the Q&A box to the left of the slides. And at that -- during that time, I'm going to run through some brief announcements. First, I'd like to thank Avantor for sponsoring BPI Digital week. Next, I'd like to quickly draw your attention to our face-to-face BioProcess International Events visiting Boston in September and Milan in October this year. Additionally, BioProcess International Europe will be delivered as a 100% virtual conference and exhibition on July 13th through the 17th. Also, be sure to check out the resource list to the right of your screen where you can download a few feature white papers. Now back to Claudia and the Q&A. So our first question, Claudia, is how far along is Avantor in transferring e-data in the biopharma space?

No, yes, great question. Thank you, Barry. So the advantage we had at Avantor, so as of -- at Avantor, we supply quite a bit into the electronics market. Let's just say we learn a lot from our EM fronts, right? In the EM space, this trend or this e-data transfer started good solid 15, 20 years ago, right? And as we have taken from that and moved into the biopharma space, right, I would say a little over the last -- over 5 years, right, we started engaging very closely with customers and defining what standards, right? And when we started in this space, the ASTM standards did not exist. So we started defining standards with customers, understanding how do we share the e-data and how do we move from the e-data sharing very importantly, into the analytics, right? So in this over little -- more over 5 years of experience, I think with tremendous amount of customer engagement, right, right? And then it has helped us with collaborating and taking a look as we have now established the e-data sharing platform for multiple customers, having us take a look, take a seat back and say, okay, what's the future look like, right? And this is what we were talking about today. What does the future look like? And how do we take it from here for the next step in terms of e-data sharing in our industry.

Our next question is how does access to data change or increase in development of COVID-19 therapies and vaccines?

Yes. The COVID world is close to all of us in the world today. So what do we do, right? And if there's been very well documented, right around -- when we develop COVID therapies and vaccines, there's quite a bit of articles up there of how AI or artificial intelligence data has been used in the drug screening, right? There's a couple of areas and therapies where we're repurposing some drugs. There's some meds in particular. There's well documented raw material variability for most of these processes, right? And as we imply in these processes, it's taking that raw material variability and making them more productive. As we think about vaccines, though, right, it's a whole wealth of information, right? There's multiple types of vaccines being thought about for COVID, and you guys are very familiar with. As we think about our more traditional vaccines, right? And in the biopharma space, recombinant vaccines these are also well documented, not fully well understood, but well-documented variability. But now as we start jumping, right, into bio vector vaccines, into DNA vaccines, right, into messenger RNA vaccines, the wealth of information that raw material variability brings at an early stage in process development for these vaccines can be really supportive as we as a world, right, try to pivot and very quickly turn into vaccine manufacturing for COVID-19. So definitely, there's a very solid argument here. As the data has been used to develop the vaccines, now pivot and utilize the richness of the raw material variability data to develop these vaccines in a more efficient way, in a more productive way to deliver across the world.

Our next question is what are the limitations of ASTM data standards?

Yes. Yes, great question, and it's -- the standard is loosely defined, right, which -- it's good and bad, right? It gives us the flexibility for those of us who are implementing and working with customers, but it also brings some challenges and one of my IT [indiscernible] colleague has most definitely taught me a lot around ASTM standards. So for example, there's required versus optional fields, right, in the field lens. And some of these are optional. And you have to collaborate with your supplier and also their customers to define what this means for us at that collaboration. The second, as you know, in building and maintaining all the cost reference for the test certificate, right? Let's say, for example, this ensures that when a CFA a test A, from a company that is sending, let's say, for example, Avantor aligns with the system of the company that it's receiving, right? So equally applies to test A and more importantly, equally accounts for unit of measure across all these tests. So while ASTM standards have helped us do a jump and have a common language, all of the fields that remain to be defined will require collaboration. And [BPOK] as an industry organization has taken a stab as -- there's 2 particular work streams getting closer to standardizing data, talking about standardizing data. So there's still a lot of work to be done collectively as an industry to get there, but there is industry forms that are supporting, and there's also emphasizing, again, collaboration with your suppliers that will help define ASTM standards.

Thank you, Claudia. Next question is what degree of validation is required of the systems that are used for data transfer? And do all data types and uses need to be fully CFR Part 11 compliant?

Yes. It's a great question. So yes, our initial focus at Avantor is all CFR Part 11 compliant raw materials, right? We're focusing on J.T.Baker materials to begin with. These are regulated J.T.Baker, right, which are done in CFR Part 11 compliant manufacturing sites. That is our initial focus. It doesn't mean that later, we would not pivot into the next product category. But that is, from an initial focus perspective, those products touch a very, very wide array of customers. So this is where we're starting. And then what degree of data validation is required in the system. There's significant amount of data validation that is required, right? So when I mentioned on the last slide, this will require specific teams both at best supplier and at the customer working on this effort, it has a lot to do with the data validations, right, and the exchanges. And so as we've gone into this with multiple customers, I can tell you this most definitely project teams at both ends, they need to work it through a project, do the data validations and systems vary, right? Multiple customers have different systems. Typically, you were having the middle of 3-party software. That helps translate the data, right, and transfer the data. But significant amount of validation will be required, thus, project teams required on both.

Okay. The next question is for the raw materials Avantor repackages and tests, will Avantor share the COA from your manufacturer?

Yes. So when we were talking about the N minus 1 raw material data, this is the area of the room that they work for. So we typically have -- as we mentioned, talked about the manufacturer and then we -- the question on the repack. This is the -- well we would consider the N minus 1, any material, particularly on the focus or the scope here of the J.T.Baker brand that comes in for a repack gets full testing incoming and full testing outgoing, right? So we are able to report incoming testing, outgoing testing. There's also stability, right? So the repack, particularly for GMP J.T.Baker products, all go through the stability system, so there is a stability testing. And then there is our CFA from our N minus 1 or our supplier, which is the area we're exploring, how do we make that accessible, right? And how could we make them accessible on an e-databases. It is data that is already digitalized in the systems, which makes it a lot easier, right? When you have data that is not digital, then that requires an additional effort. So yes, most definitely, that's an area we're looking into.

Thank you. The next question is, what is the most influential change you have implemented recently to increase raw material quality?

Well this is a continuous, continuous improvement, right? And with the focus on continued improvement, raw material quality and it's a journey, right? It's a journey that we continue to travel. And hopefully, all of -- everyone around the industry does in the supply, right? So from the raw material quality, the most latest is, as you guys know, Avantor and VWR came together as a company a little over 2 years ago now. So as we think of raw material quality, we're focusing a lot on raw material standards, right? How do we make sure that as a company, as we went through integration at multiple sites and multiple product lines, we have one raw material standard that we can very clearly communicate to our customers, this is an Avantor quality standard, right? There -- as we -- as the acquisition went through an integration, right, you can imagine breadth of product lines, rest of facilities. And the key focus for us is that standardization of the raw material quality and quality standards, right, and making sure that any product you get from any Avantor facility, be it legacy, VWR legacy, Avantor comes in with the same quality standards and raw material quality.

The next question is do we use different vendor raw material while process development for the manufacturing? And if yes, how many different vendors should we consider or use during process development?

Yes, for sure. So it's a very good question, right? So typically, as you think around your process development matrix in your process, it will depend a lot on your process, right, and what your process requires. So if you have a process that you think is going to be very testy of certain, let's say, upstream supplements, right? I would highly suggest that you consider making sure you get the highest purity and the -- considering suppliers that really assess the raw material variability. And well-characterized materials, right? And what I mean by well characterized is there's -- as the suppliers, there's a range of tests you can do, right? And just to be very, very long. And as you think about which steps in the processes raw materials go into, I would consider selecting suppliers that have very well-characterized raw materials. This would help most definitely in your process development, right? And I -- typically, from a risk perspective, you would typically be thinking about different vendors. And to the question, yes, you'd be using different vendors of raw materials. The key is, I think, well-characterized in high-purity materials first? And as you think about once you get there, right, are there 2 vendors, right? Can you risk and mitigate as you do your process development, definitely wise to do? But think about what does your process require, right? And how comfortable do you feel with levels of variability there are -- shared with you from your supplier? And that would be a good starting point.

Thank you, Claudia, and thank you for such a great session. You've answered all the questions that have been submitted at this time. I want to let our attendees know that this session was recorded, and you'll receive a notification in 24 hours when the on-demand session is available for viewing. But before you log off, please take a moment to complete the feedback form so we can continue to improve your digital week experience. And on behalf of Informa Connect Life Sciences, thank you for joining us, and I hope you have a great day.

Thank you, Barry. Thank you, everyone, for your time today.