Subscribe to our newsletter and stay informed

Check out our list of top companies

Check out our carefully compiled lists of the most relevant and impactful companies within their fields.

Check out our list of top unicorns

Read and learn about the biggest companies that various countries have produced, how they made it, and what the future looks like for them.
December 22, 2023

A Material World: An Interview with MaterialsZone CTO Ori Yudilevich

How the materials informatics company is changing R&D with one single platform

Materials informatics is a field that melds the power of big data and machine learning with materials science to reduce costs, improve products, and send them to market sooner. We spoke to Ori Yudilevich, the chief technology officer of MaterialsZone, a leader in the materials informatics landscape. Combining technical expertise and science, MaterialsZone has been instrumental in developing cutting-edge platforms that are setting new standards in the industry.

The firm is having a transformative impact on materials informatics in various industries. We discuss how this drives efficiency and innovation and contributes to a more sustainable future.

We also get insights on the future trajectory of materials informatics, the challenges of bridging AI with materials science, and the impacts of these advancements.

MaterialsZone CTO Ori Yudilevich

Neil: How does machine learning help collect data for the materials research and innovation process?

Ori: MaterialsZone has a materials informatics platform that serves industries that create material-based products. So that's products where some kind of materials research is involved. The development of new products is kind of like a cooking process. There's kind of an iterative process where you try out different materials, you try out different processes in cooking that would be trying out different cooking times at different temperatures. Traditionally, this process is very slow, and it requires a lot of trial and error.

Today, with machine learning, you can really increase the speed of this process, and you can simulate a lot of things that you would otherwise have to do in the lab and that can take a lot of resources and time. This machine learning process can actually recommend what you can do. It can try out a thousand or a million different things that you would otherwise try in the lab, and it will tell you these are the top ten, why don't you try these out?

Then the whole data collection, which takes place in the lab becomes much faster and much more efficient.

Neil: How much of a negative impact does slow or inefficient data management have on the innovation process?

Ori: This relates to the story I just told because the original process before machine learning, before digitalization, was basically trial and error and using scientific intuition, which is good. We’ve progressed quite a lot in that way, but still, it was a very slow process, it could be years of work until we can get to the product that we would like to get to.

Because the process was very long, there was also a problem of knowledge loss. So, in many years, a lot of engineers and people who have a lot of knowledge would leave the organization, new people would come in, and there would be duplicate work.

Digital data management allows you to keep this data alive and in one place. It allows searching for old data, learning from what people did in the past, and today, with machine learning, a lot of this learning is done automatically, and this empowers the engineer in the lab. These technologies allowed for a lot of improvement.

Neil: Working in a lot of different industries, how do you tailor the platform to meet the requirements of different industries, especially if it's a niche area?

Ori: That’s a good question. We do have a lot of issues because every time we discover domains, new domains have new challenges. However, we do see that the core of what we do, the core of our platform, is more or less common. So, as I said, it’s a bit like cooking. It doesn’t matter which domain you go to, there is this iterative process that I described before, and that is common. The core of our platform actually fits all industries, more or less.

The parts of the platform that may be domain specific are the parts related to how the data is being input. Each domain has its own kind of machines that are used for collecting data, for making measurements. In addition to that, each kind of domain has its specialized calculations and analyses that are being performed. So it is really the connectors, the way we input the data.

It's often about integrations, being able to integrate different systems and also about adding modules that allow us to do these specialized calculations. These are often either developed by us or our clients, who can do the calculations on their side, then, using our API, upload the data to a more generic platform that can serve them well in a unified way.

Neil: Are there any features or tools that support individual industry innovation and market responsiveness?

Ori: Yes. To give you a few examples, we work with the battery industry. We have to measure things like current and voltage. We work with the building materials industry, where you have to work with the actual strength and different properties of cement and other building materials. Each one is a completely different domain, a different academic degree you would do maybe to work in that domain, and different mathematics.

We have the ability to create something we call analysis apps. These are kind of little calculators that could be interactive or automated depending on the situation,. Customers also have the ability to build these themselves. We can provide a developer kit, if the company we work with has in-house data scientists. That's a whole feature set that we offer to adapt the platform to specific domains.

Neil: Can you describe for me the security measures that you have in place to protect sensitive data or intellectual property?

Ori: Security has to be placed as the highest priority. You see a lot of platforms being attacked or data being leaked. We have a security team that deals with this. We are now at the final stages of getting a SOC 2 security compliance which basically covers the whole spectrum of security, availability, confidentiality, and privacy of data. This means that we put all the walls around the sensitive information, like encryption and firewalls, in order to protect our customer's data.

In addition to all those security efforts, our platform also has security built in. If we have a big organization that we’ve onboarded, these organizations might also have a need for permissions between divisions. Maybe not all divisions are allowed to see everything. Maybe you want some divisions to have only read-only access to specific types of data. We've developed an extensive granular permission system that allows us to allocate permissions at different levels and for different entities in the system.

Neil: How has a background in data science and software development Ops helped you in this role?

Ori: It helped me a lot. Of course, there were a lot of things that I needed to learn. Throughout life, you learn certain things, and then you start a new job, and you see that you need other skills as well. I think I have a pretty diverse background because, from my childhood, I've been programming and had several jobs where I was a software developer, also a product manager in the early stage, but in addition to that, I also did a detour and came back into academia where I spent many years studying pure science. I studied physics, chemistry and mathematics.

I think the combination of these two helps me a lot because our product is a scientific product. It serves the scientific community. Having been part of that community I understand our clients better. I find it easier to understand their way of thinking.

On the other hand, in the end, we are a software company, building software. I think this combination has aided me a lot.

Neil: In what ways does your platform contribute to sustainability and eco-friendly practices within the materials science industry?

Ori: It's a big issue these days. We're getting into it more and more. I think it helps in two main ways. One way is that we facilitate the calculations and reporting that needs to be done on sustainability. For example, companies today that create products made out of materials, have to report on the carbon footprint of the product, and our platform allows them to take into account all of these things. Collecting all that data and facilitating the calculations that they have to do and the reporting, that's one way. That relates more to the sustainability division of a company, if it's a big company, or if it's a small company, maybe even the R&D would do that type of work.

The other thing is actually allowing our clients to take into account the carbon footprint and the other sustainability parameters in the development of the product. It becomes like another parameter. In the past, you would just say I want to have the most efficient photovoltaic cell, but this could mean that you would use materials that have a high carbon footprint. There sometimes might be competing parameters. Maybe something that would give you more efficiency might actually also be less sustainable. You optimize these two, it's a multi-parameter optimization problem and our platform allows them to take into account all of these.

In addition to sustainability, there are also other regulatory constraints. For example, if you're making a plastic product for the food industry, there are certain raw materials you're not allowed to use, or you're allowed to use very little of them because they could be poisonous in a product that touches food. Our platform allows all of these parameters to be taken into account in the R&D process.

Neil: Do you have any specific examples of how your company has aided another company that perhaps would not have been possible without the platform?

Ori: We work with some very large organizations. We also work with startups and medium-sized companies, and each type of company has its own set of challenges. In the case of big companies, there are a lot of traditional companies that were not working on digitizing their systems in the past. Also, throughout the years, they've accumulated a lot of systems that you see in large companies, especially if you look at companies that buy other companies. So, each new company they buy comes with their own set of data systems. Could be CRMs, R&D-related systems, or inventory management, a whole array of systems and integrating these becomes very challenging. It becomes very difficult to take the whole flow into account.

We see ourselves as an end-to-end platform.

It's a platform that connects all the pieces because everything is important in the end for producing the best product. So data from procurement, you might want to take into account the prices of the raw materials, sustainability, and regulation, and the R&D itself. We have a lot of data itself to take into account with things that you’re measuring, the process of developing the product, all the way to quality control, the actual manufacturing line, all of the parameters you also collect there go back to R&D,it's like a feedback system.

Our platform allows for connecting all the pieces, and better decision-making at all levels. R&D can take into account all of these things that I said. Procurement has much more visibility into how the raw materials that they purchase are being used in a later part of the process inside the company. This is one big focus of our product.

Neil: Can you share any future updates or developments we may see on the platform in the future?

Ori: In the end, we’re a SaaS platform. That means we're hosted on the cloud. You don't have to install anything on your own servers, and one of the nice things about SaAS is, all these platforms we use, like WhatsApp, and Gmail, is that they keep improving all the time. Every morning we wake up, we find a new button that does something cool,after having learned from how users are using the system, where they're getting stuck. So we're constantly improving the user experience and increasing the usability, and putting in cool new features depending on what we see that our users need. This is one aspect, just ongoing improvement of the platform. That sounds maybe less exciting, but I think that is where the big value is.

In addition to that, we’re working on two big projects; one of them is this whole machine learning iterative process that I was discussing at the beginning, how machine learning can learn from the experiments in the lab and suggest the next experiments you should do. So introducing some new algorithms and technologies that are pretty modern and are only being used in academia at the moment into our platform, and making them available for industry.

Another thing we’re working hard on, just like a lot of other companies in different areas, is understanding how we can leverage generative AI, tools like ChatGPT in our platform. We’ve already identified several places where this amazing technology can solve problems. Up until now, you needed maybe to hire one hundred students to look into a lot of text and extract information, which is very useful, public information, For example, specifications of raw materials or information from academic articles. All these things are publicly available but very difficult to put into a structure that can actually be utilized. This is just one example of how generative AI can help a platform like ours, and we’re working hard now to build something. We always start with a POC and check if it does give value and then improve it and make it part of our product.

Neil: How do you foresee the evolution of AI in materials mathematics in the next five or ten years?

Ori: Different industries are at different places in this process of digitalization. Maybe the pharma industry is more advanced than the whole industry of material-based products. We see that a lot of the companies we work with are trying very hard to catch up. But they're still not at the same level as other industries, whereas this industry like many other industries, are a huge part of our lives. Every physical thing that is in your house, is actually a product of this industry, could be a table, your laptop, or solar panels on the roof.

We’re living now in this revolution. I've been in the company for four years. The company was founded six years ago, and in these six years, we've seen huge evolution, and significant change. Six years ago, we had to tell people what is materials informatics, what are the benefits of digitalizing. Today, companies come to us. You don't need to explain this. We’re at a different level already. People are much more open to making these changes.

Once companies are digitized, once all the data they have is not siloed, and it's all combined or connected in a way that can be leveraged for applying these AI tools, the pace of innovation is going to increase exponentially. Given the state of our world, with global warming, we have to see much faster innovation, especially in areas like alternative energy and biodegradable materials. As we've seen in the past centuries, tech always catches up with the needs of society. I think we’re seeing that happening.

Neil Hodgson Coyle
Neil Hodgson-Coyle
Editorial chief at TechNews180
Back to top

Related articles

chevron-down linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram