4.28.23-Korea-Kebotix

ASHISH KULKARNI: I hope you can understand my accent. Just so you know I'm speaking English.

[LAUGHTER]

So I'm here to talk about AI/ML. I represent Kebotix. What we do is new molecules for chemistry. If you are thinking about formulation, if you are thinking about reaction, if you are thinking about synthesis, that's who we are.

We are not an API. We are not a drug substance company. We are a drug product company. Our co-founder is from MIT. We have quite a few employees from MIT that we have in our location.

If you think about our team, it's a combination of people who know AI/ML on the left-hand side, and on the right-hand side are people who have hundreds of years of experience in chemical industry. We try to combine the two and help people think about synthesis. If you remember Jim Collins' talk, he talked about how synthesis sometimes requires more optionality. That's what we do.

If you think about chemical industry, and here I'm talking about chemicals outside of APIs, outside of drug substance, there is still a lot of innovation needed there. It's full of friction. There are a lot of regulatory challenges, innovation challenges. On left-hand side, you will see there are tons of business-related issues, talent-related issues.

On the right-hand side, there are technical issues on how you come up with a new molecule, how you do manufacturing, and how you do customer qualification. And with Kebotix, we have come up with a model that uses AI/ML to design new molecules. We have curated about 22 databases, 114 million molecules, to think about how to design a molecule for a given set of properties.

Once we design the molecule, we also can make the molecule. We have curated about 7 million reactions on how to synthesize that particular molecule. And we also have testing capability. And we also have sampling capability with our partners.

The differentiation that Kebotix has got is, of course, it can work with a wide variety of data. It can work with customer data or external data, a wide selection of predictive algorithms, and also a significant amount of infrastructure capability so your data can stay secure and safe. That's what the differentiation is.

Here is an example. I don't have time. But if you stop by at my table, I can walk you through it. It's an example of a pigment. But it represents a small molecule, where we went from an idea to a sample, a physical sample, in less than six months. On the right-hand side, you will roughly see our process.

We started with about 4.3 million molecules to get a certain set of properties. By following the process, we generated another 100 million molecules, basically reinforced our algorithm and gave five samples to a customer to test it out. That's what the company is capable of.

Here are all the properties we can predict. We can predict physical and chemical properties. We can predict environmental properties. We can predict quite a significant number of light and spectral properties as well.

And we are continuing to add to this predictive capability, as we continue to build the company. If you are in the audience today as a pharma company, we can do the following. If you are in small molecule, we can work with you in terms of synthesis or any kind of green chemistry or any kind of yield optimization projects. That's something we know how to do.

If you are in biopharma, we can talk to you about CMC. We can talk to you about excipient designs. We can talk to you about solvents that you don't like to play with in your operations, or if you want any green chemistry to go with it. If you are in medical devices, obviously we can think about materials, colors, polymers, and that kind of formulation optimization.

If you are in new modalities, we can talk to you about reagents and how to innovate in that area. In terms of the capability, we also have a SaaS model to go with it. If you don't want to work with us, if you just want our capability, we can also give you that at very, very measurable and very economically viable SaaS capability. We do it both as a service or we can also do it as a product for you, depending on which business model you are comfortable with as a company.

Here are some examples, just in terms of data point. We have developed over 200 new molecules, in terms of doing it for customers on demand, on an average cycle time of about three months. We have quite a few customer paid projects.

On the right-hand side, you will see eight examples where customers were either able to pattern their own molecule or we were able to pattern the molecule based on what they were looking for. That's what the track record is. Though it's a startup, there is quite a bit of proven capability.

In terms of my ask, I'm looking for customers to work with us in pharma, biopharma, materials, or chemicals. We can also work with you on downstream capabilities. We are also looking for collaborations. If you are interested in looking at your CMC operation, if you are looking at how to do synthesis and reaction optimization, or if you are looking for some green chemistry capabilities, but all driven by reactions and chemistry.

If you have some new challenges that I have not thought about, or we have not presented here, very open to working with you, whether it is optimization, whether it is data workflow. If you are sitting on a lot of data and you want to think about organizing data, we can work with you on that as well.

I'll be outside, like all of us. I'll be talking about the same things. If you are interested, please stop by. We'd love to hear your challenges and see how we can help. Thank you.