Building Rhea's Factory: How AI-Designed Enzymes Could Finally Solve Plastic Recycling

Summary

---

Just Now Possible — Episode Summary

Podcast & Episode Details

• Podcast: Just Now Possible
• Host: Teresa Torres
• Episode: Building Rhea's Factory: How AI-Designed Enzymes Could Finally Solve Plastic Recycling
• Date: May 14, 2026

---

Episode Overview

This episode explores how Rhea's Factory is marrying synthetic biology and AI to develop next-generation technologies for recycling plastics at the molecular level—going well beyond traditional recycling to enable genuine circular economies. Teresa Torres, the host, speaks with Arzu (CEO & co-founder, molecular biologist) and Mert (CTO & co-founder, engineer), unpacking how their deeply interdisciplinary approach is tackling the longstanding problem of plastic waste using AI-driven enzyme design.

---

Key Discussion Points & Insights

1. Origins & Mission of Rhea’s Factory

• Founders: Arzu (biology/biochemistry) and Mert (engineering, ex-Google PM) joined forces to attack plastic waste by fundamentally rethinking recycling (00:09, 00:33).
• Vision: "Our inspiration is coming from...we have all the materials out in the world right now. We don't need to dig oil to generate new materials...Let's use the waste and generate value from them. And our tool is biology to be able to do that." (Arzu, 02:13)

2. The Problem with Standard Plastic Recycling

• Traditional Recycling Limitations:
  • Current methods are mechanical or chemical, largely just melting and remolding plastic (03:48).
  • Quality degrades with each recycle; plastics can typically only be recycled 2–3 times before becoming waste (03:48).
  • "You simply melt the form and then reform it to the new shape. That...degrades the quality of the polymers..." (Arzu, 07:52)
  • Only about 10% of plastics are recycled globally (05:01).

3. Biological Approach: Enzymes as Game-Changers

• Why Use Biology?
  • Biology offers selectivity and specificity: enzymes can target specific bonds in mixed waste streams (05:57).
  • "Life can happen in a very small environment within the cells with millions...of different reactions...Chemistry can only do one reaction at a time and cannot generate selectivity." (Arzu, 05:57)
• How Their Process Works:
  • Enzymes break plastic polymers down to their original building blocks—monomers—essentially rewinding them to the beginning of their lifecycle (09:03, 11:18).
  • This process is low-energy (60–65°C) and at atmospheric pressure, avoiding harmful byproducts or emissions (13:31, 17:11).

4. Scientific Foundations: Protein Structure & AI

• Protein Folding:
  • Enzymes are proteins; their function is determined by their 3D structure, which was historically very difficult to predict from their amino acid sequence (23:28–26:11).
  • AI and tools like AlphaFold now predict protein structures, vastly speeding up discovery (24:05, 24:58).
  • "[Previously] it was taking a PhD student lifetime to discover the structure protein...Now with AI technologies...the linear structure can be given and what it will look like in the folded state could be predicted." (Arzu, 25:50, 24:58)
• Implication for Enzyme Discovery:
  • Enables rapid and novel enzyme design, expanding beyond what nature has evolved (29:14, 61:15).

5. The AI-Driven Enzyme Design Workflow

a. Problem Framing & Workflow Structure (41:01)

• The process starts with a problem statement: e.g., "We want to break this polymer."
• Inputs can be enzymes found in nature but need enhanced properties for industrial use (45:08).

b. AI/ML Models & Pipeline (37:17–39:38)

• Protein Language Models (PLMs): Used as foundational models, analogous to LLMs for natural language (34:37, 37:17).

• Multi-step Workflow:

  1. Generation: AI generates candidate enzyme (protein) sequences (37:17).
  2. Structure Prediction: Predicts enzyme folding and structure.
  3. Function Prediction: Evaluates likelihood that the structure performs the required catalytic function.
  4. Wet Lab Prediction: Assesses feasibility and experimental success likelihood.

• Orchestration:

  • Initially human (“human orchestrator”); transitioning to agentic orchestration by AI (44:08).
  • Guardrails are present at every step to ensure candidates meet design criteria (46:34–47:56).

c. Data & Feedback Loop

• Model is trained and improved with data from both literature and private wet-lab experiments (43:18, 48:03).
• Every laboratory run that succeeds or fails refines the model, improving future predictions and "hit rates" (48:43, 49:19).

d. Diversity & Exploration vs. Exploitation

• AI enables broader and faster exploration of the enzyme design space than nature or manual methods, leading to potentially revolutionary solutions (60:57, 61:09).
• Sometimes, "hallucination" (creativity) in the language model is desirable: pushing beyond known structures to novel functional enzymes (63:57).

6. Transition to End-to-End, Autonomous AI Platforms

• Current focus is end-to-end process optimization, not just enzyme design but also downstream parameters (64:31).
• Envisioning a future where even wet lab testing is minimized or replaced by advanced prediction models (65:33).
• "Our platform's customer is our lab and that our lab's customer is end to end process." (Mert, 41:01)

7. Industrial Scale & Impact

• They're piloting a 5,000-ton-per-year demo plant in California, partnering with a bottling company (65:38).
• Long-term vision: custom enzyme and process design per customer/product—bespoke, efficient recycling tailored to complex real-world waste streams (67:05).

---

Notable Quotes & Memorable Moments

• On the paradigm shift:

  "We have all the materials out in the world right now. We don't need to dig oil to generate new materials...Let's use the waste and generate value from them." (Arzu, 02:13)

• On how biology wins:

  "Chemistry can only do one reaction at a time and cannot generate selectivity...Our catalyst enzymes...are evolved for years in nature to be very selective..." (Arzu, 05:57)

• On AI’s revolutionary role:

  "Now we tackle most of this problem computationally, using machine learning tools, and design new enzymes that nature hasn't seen before." (Arzu, 29:27)

• On the scale of discovery:

  "Nature has evolved to that much...design space. Now we are able to explore all the other pieces that nature hasn't evolved into. And that is...mind blowing." (Arzu, 61:15)

• On the value of AI ‘hallucination’:

  "In some applications or some modules, I want it to hallucinate...We want to explore that design space." (Mert, 63:46)

• On clamshell plastics:

  "For us, this material and the water bottle, they are all the same...We break them down all the way to the beginning." (Arzu, 68:13)

---

Timestamps for Important Segments

• Intro & Founder Backgrounds: 00:09–01:51
• The Core Problem in Recycling: 02:13–05:57
• How Enzymes Enable True Recycling: 07:11–13:28
• Protein Folding Breakthroughs & Impact of AI: 23:28–26:35
• Detailed Workflow of AI-Driven Enzyme Discovery: 34:37–44:08
• Laboratory Testing, Private Feedback Loop: 48:03–50:12
• Expanding the Search Space and Embracing Hallucination: 61:09–64:15
• Scaling Up, Customization of Enzyme Design: 65:33–67:57
• Real-world Application (Clamshells Example): 68:11–69:33

---

Tone & Style

The conversation was enthusiastic, technical, and collaborative—with Arzu providing scientific depth and Mert translating computational complexity into product context. Teresa’s tone is curious, supportive, and clear, teasing out both the big picture and technical specifics for listeners building in the AI space.

---

Conclusion

Rhea’s Factory is pioneering a new model for tackling global plastic waste—leveraging AI as an “enzyme scientist” to open up vast new design and recycling possibilities, speeding up the transition to a circular, sustainable materials economy. Their approach also exemplifies a broader shift: how industry-specific agentic workflows, built atop powerful scientific foundational models, can bridge real-world needs with cutting-edge computational biology.

---

Transcript

[A]

Welcome to Just Now Possible with Teresa Torres.

[B]

I'm Arzu, co founder and CEO of RIAS Factory. I'm a molecular biologist and biochemist by training. Had long years of academic training, eventually find my passion in synthetic biology and circularity and founded REAS Factory two years ago together with Mert. Mert.

[C]

I am Mert, CTO and co founder of REAS Factory. I have been in the technology world for 20 plus years. I'm originally from Turkey, came to United States Bay Area because I'm a nerd, worked at big companies and before REAS Factory I was at Google for 10 years. Early on project manager and then became a product manager there. And I wanted to be able to use my skills for something in Pathfind. That's when I met Arzu and her vision about changing disrupting plastic recycling with AI was like really meaningful to me. So that's how we join our forces.

[A]

Yeah, yeah. I'm really excited to dig into what you're doing because it sounds very meaningful. Arzu, it's clear you have a biology background and a biochemistry background. Mert, do you have a biology or biochemistry background or are you primarily bringing the engineering mindset?

[C]

No biology background at all? Biology was probably my worst class in my years. I think I learned a lot the last two years thanks to my good teacher Arzu here. But I'm definitely coming to purely from computational perspective.

[A]

Excellent. I like this. This is a good, very different domain expertise but clearly found a way to work well together.

[B]

Indeed.

[A]

Okay, so for Rhea's factory it sounds like Arzu, you had the seed of inspiration. So tell me a little bit about what does Reyes factory do and where did this seed of inspiration come from?

[B]

We develop technologies for recycling materials and specifically starting with plastics and we use biology to do and our inspiration is coming from the fact that we, I believe and we both believe we have all the materials and are out on the world right now. We don't need to dig to generate to we don't need to dig oil to generate new materials. We make new materials and we waste them and we throw them away. And that linearity of our manufacturing and the whole world actually bothered us a lot. This is where everything started. It's why don't we use the materials we generate, once very valuable and eventually ending up in landfills? We why don't they become valuable and that is actually the core of circularity and that is where our company is started. Let's use the waste and generate value from them. And our tool is biology to be able to do that.

[A]

Yeah, I like this. Okay. I know from your application, you framed this as. Today we already have technology for recycling plastic. If I understood correctly, we can only do that so many times. Each time we use a recycled plastic, it degrades more and more. Maybe we can recycle it two or three times. Is that accurate?

[B]

That's correct. The definitions are actually important. Currently, the way we call recycling is mostly reusing the materials. Meaning we collect the materials, we melt them down, and reform them into new materials. That causes the quality to degrade by time. Therefore, you cannot really do this so many times. However, in our case, and what we do is we are really turning the materials back to their original building blocks so that it is like, to the beginning of their lifetime and they can start being a new material again with the right process applied to them.

[A]

Yeah. So I want to get a little bit into the science enough to give context for when we're getting into what your agents are actually doing. So I would love to hear, like, with traditional recycling, like, what's happening at the, like, maybe high level, molecular level to understand, like, where that waste is coming from, and then maybe talk a little bit about the approach you're pursuing and why that doesn't have the same waste.

[C]

We have a plastic problems. Everyone accepts this.

[A]

Yeah.

[C]

And we can only recycle 10% of the plastic we manufacture. And it's been like a hundred years. We are trying to solve this problem. Like, why? Yeah, the reason, one of the reason is that petrochemical companies are manufacturing plastic. They understand mechanical technologies, they understand chemical technologies, and they were trying to solve this plastic recycling problem with the tools that they understand, which is mechanical and chemical. Unfortunately, those technologies have an inherent limitation to make plastic recycling 100%. That is why I was pulled into her vision, which is we need a new tool here, which is biology, because we have been trying to solve this problem with the same tools. We created the problem.

[B]

The biggest point about biology in this space, as Matt said, we are bringing biology to the chemical industry. And what biology brings, actually is selectivity and specificity. Life can happen in a very small environment within the cells, and there are millions, tens of thousands of different reactions happening in a very small space. Chemistry can only do one reaction at a time and cannot generate selectivity because the processes used are either the heat or high pressures. Very general. But our catalyst enzymes, the biological catalyst, are evolved for years in nature to be very selective to different materials and different processes. And that brings a huge potential into the chemical industry. Meaning the real problem in Recycling currently is the waste being mixed mixed with different materials.

[A]

Okay.

[B]

And that is really difficult to tackle with traditional methods like mechanical or chemical. But biology can tackle it with different catalysts designed for different materials.

[A]

Yeah. Okay. Help me understand. Arzu, you described initially like what the chemical companies are doing is they're literally melting the plastic, which whether that's a chemical process or literally a heat process, like I'm curious about what is getting lost in that process. And then talk to me about how using targeted enzymes, what gets preserved with

[B]

your solution with the mechanical methods and applying heat, you actually do not turn them back to their molecular building bloods.

[A]

Okay.

[B]

Just simply melt the form and then reform it to the new shape. That what it does is it degrades the quality of the polymers, the. The building blocks of the material. And which doesn't lead you to generate the quality that is being needed for in the industry. What we do is maybe I give this example that will help Heather explain. Materials are made out of polymers. You can think of them as necklace of pearls. And what we do is we break this necklace into the single pearl pieces and then rebread them together. But the traditional method simply put that together, melt and remold it. And doesn't really achieve the quality that is needed.

[C]

Yeah.

[A]

Help me.

[C]

The raw materials as an output from our process recycling process are plug and play to the existing supply chain as if they are coming from wherever, oil, et cetera, like as high quality as

[A]

they can be in that melting process. Are you losing polymers? Are the structure of the polymers changing?

[C]

Yeah.

[B]

So with the traditional methods, what is critical is the length of these necklaces. And she has. And the traditional methods just break down this chain length and make them shorter. But they don't break them down to the building blocks, they just degrade it. And what we do is we really bring it down to the original building blocks by eating down the polymers and cutting precisely at the right locations and generate the building blocks.

[C]

Enzyme is part of the solution, but it's not the only part.

[A]

Okay.

[C]

So when we talk about, when we realize that, okay, we design an enzyme, but then we need to have the end to end process and the enzyme is a depolymerization process. All the activities happening in that process, the material we add, the base that she selects to add comes into play. That's where we are starting to use AI for example, to optimize not just the enzyme design for the sake of enzyme design, but also how that enzyme will operate in a more unit, economic feasible way in the full process.

[A]

Okay. So if I understand right so far, when a traditional company is applying heat to the plastic, we're shortening polymers, there's some waste. Help me understand why enzymes change that. What is it? I know enzymes are more targeted. Help me understand like what's happening. We're not getting that polymer shortening. Like what is it about enzymes that allow this to happen.

[B]

What happens is so the original process, where it starts, maybe this is what we talk so that it makes more sense what enzymes do in the life. It's more like going back in time of the life of a polymer. So the life of a polymer starts from what is called the monomer, the building block. And then there's a process to generate these polymers and the polymerization. And then you make different materials from these polymers. There are additional additives added during this process as well, which is not that important right now. But what happens is traditional methods take the final product and turn back them to the polymers. Again, not more than that, but these polymers are not the polymers that are ideal sized, but they are smaller now because of the heat. And what enzymes do is they really turn the timeline back all the way back to the monomers, meaning that they are now original building blocks, the time zero for the life of a polymer where they are just monomer. And you can generate as long polymers as you like with your process.

[A]

Okay.

[B]

That is why the new materials being made from the enzymatic process can be brand new quality. It is like equal quality as the chemicals you make from oil eventually.

[A]

Okay, so like the first time plastic is being made, first monomers are made, they get combined into becoming polymers. When we recycle, we're not traditionally, we're not getting all the way back to monomers, we're getting to shortened polymers.

[B]

Exactly.

[A]

And what you're doing with enzymes is you're trying to deconstruct the polymers back to their base monomers so that like for a plastics company, you could input either one and they're going to look the same.

[B]

Exactly. And that's the value that we bring.

[C]

Enzymes are basically a byproduct from a bacteria and they are living organisms. We are not using living organisms in our process. We are not using bacteria, so it's make it scale even easier. However, they are biological organisms. Therefore they live and operate in a very mild conditions. Compared to 1000 Celsius degree heat that we process, we are like operating at 60 Celsius, for example, which means that our energy consumption is much less. Our greenhouse gas emissions are much, much lower. Because when you apply Heat to this process. Then you can also generate hazardous materials or hazardous gas. Our process is the most sustainable in terms of plastic recycling in the existing technologies.

[A]

Okay, so I'm going to ask a really naive question, so humor me. I'm familiar with how enzymes work in living organisms. Like at a very rough level of. They catalyze things normally. In plastics, there are no enzymes. Correct. Plastics are not living things. There's no. Okay, so really what it is, you're using enzymes in a process to deconstruct plastic into its basic units.

[B]

Correct. Okay, take the waste and I will show you this. Like, I. We just have some blueberries here. I'm pretty sure you have. We take this when it is becoming the waste, put it in our reactors together with our enzymes to break this material back to real building blocks. And I can actually show that building blocks as well.

[A]

And then Mert, you said something that, like, intuitively makes a lot of sense to me. If we're melting something at 1,000 degrees, you're going to get a lot of byproduct waste. There's going to be disruption that you don't want. And I'm assuming what's happening with your enzymes is they're deconstructing in a much more targeted way and that's what's allowing these building blocks to stay intact.

[B]

Correct.

[C]

And Arzu mentioned that they are selective. So if I give enzymes with this plastic that I want to recycle, maybe I have a T shirt from plastic polyester, but it is also cotton. It has something else. Our enzymes will find the polyester pet the targeted plastic type, for example, and will only catalyze or break down into monomers. That one. The rest is left over. But when you apply heat. Doesn't separate.

[A]

Yeah, it just burns it all up.

[C]

Exactly.

[A]

Yeah. Okay, help me understand. And again, maybe some of this is secret sauce, so that's okay. You can share, like at a high level. But like you're adding. You said you're a reactor. You're adding enzymes to a plastic. Like, I guess in a lot like a living system, I can see how like an enzyme acts in a cell. What is your process that you're creating this like, biological process? I'm imagining a giant petri dish.

[B]

No, this is actually a reactor.

[A]

Okay.

[B]

Reactor means a closed container where we are able to mix the content so that the reaction can take place more, much more efficiently and we are able to apply heat to this container. And that's all. And the heat we are applying is, as man said, is 65 centigrade degrees compared to hundreds to thousand centigrade degrees in the other contexts.

[A]

Okay.

[B]

And more importantly, the pressure is also atmospheric pressure, meaning we don't apply any additional pressure.

[A]

Interesting.

[B]

It is happening like in your normal kitchen where like something is being mixed and processed.

[A]

Okay. So everything goes.

[C]

When I saw this process, I was like fascinated because it's so simple. Yeah. We just put water, plastic enzymes and that's it. And he did 265.

[A]

Yeah. Arzu, I'm curious about where the seed for this process came from. I know there's a lot of like creative solutions in science come from like somebody from another domain looking at a problem. And it sounds like this might have be what happened here. You have a biology background, you got curious about recycling. Traditional recycling has taken a very chemical physical process. Tell me a little bit about just that seeded discovery for you.

[B]

I will go back, I will go to a place beyond myself. So this idea and concept is nothing new to science. Science is always looking for new inventions. And also understanding life takes you, helps you to translate it to the other fields. What happened is scientists have been working on this problem making using the existing feedstocks, meaning the living feedstocks in a much more useful manner. And this has been going on, but specifically for plastic. There had been no effect to enzyme found so far until scientists from Japan actually made a discovery from the remaining of a recycling facility. It was a bottling recycling facility and they were collecting samples of living bacteria within that remains of recycling facility. And they discovered a bacteria that has the potential to break down pet the material that I showed you here and that was groundbreaking because it was the first time that a living cell was able to live on plastic. But of course this wasn't something that could immediately be applied in an industry. It just opened up a space in science. Hey, it is possible life has evolved in a direction. If we put the life in a context where there's a lot of this material, then life tries to go and find a way to make use of that material. Eventually the enzymes responsible for that process have been identified and scientists start working on that. And this is where I discovered that. I became aware, hey, there are these enzymes. It's so cool. Everyone finds the scientists find this really cool. But then when I go back to real industry, I didn't see anything happening in that area. And that was the frustration. It's like, why no one is taking this. It's such a cool idea, right? Like, why is it not happening? And that is the beginning of it. Like, hey, I Think it really touches something important, but it hasn't yet been applied. And initially I start understanding why is the case. So my story actually start by quitting my previous job, which was again for a climate technology startup. But I wanted to dedicate myself to this problem and literally spend maybe six months exploring what is happening in the recycling space right now, why no one wanted to implement this technology. I didn't even know what it takes a new technology to be implemented in the market at that time. The story started that way, the background science done by other scientists. But I myself had been working on different enzymes and engineering them for industrial applications. That is what I did in my previous jobs and focused on doing that. That is, hey, there is another enzyme out there that can really help another industry to change the path. But no one was doing it. But then when I started exploring, I start seeing, oh, there are actually a couple of others start doing that, but they are not yet at the market. And seeing at the same time, what was also happening is AI was changing the enzyme discovery field and it was changing it in a very dramatic way. This was eventually getting a Nobel Prize in a very different context. I was like, I think there's no reason not to get into it now. Hey, we can make these enzymes that nature has evolved, make them even better, apply them in industry and go for it. And I thought maybe naively, like no one else than we will go and do it because they don't know what the enzymes will do, that what they will do. And that's how the story started and that's how we joined forces with MERTA as well, because we both wanted to do something impactful. That is the core of what we are doing, the real foundation. And I wanted to apply a new technology in engineering biology and applying it to a chemical industry. And MERT was completely on board with the idea and that is how it started.

[A]

Yeah. Okay, so it sounds like you already had a background in enzyme design applied to industry, just not this industry.

[B]

Correct.

[A]

And then it sounds like a second, I'm sensitive to the fact that I'm going to say catalyst was the Nobel Prize and if I remember right, that was like in protein folding, is that right?

[B]

Correct. Correct.

[A]

Yeah. Okay, so not exactly the same thing. Do you want to talk a little bit about the differences there?

[B]

Yeah, it's a terminology thing. Protein enzymes are proteins as well. Proteins are one type of building blocks in life. And what is special about them as well? They are actually linear chains, but they function as folded structures. Yeah. And that is the key here And I had been doing my PhD also on this problem, how proteins are being folded to a certain structure and not to the other one. And how this process happens.

[A]

And this structure, if I remember right, impacts its function.

[B]

Correct. And the thing is, the same chain will never fall to a different structure. It will always fall to the same structure. Same exact chain.

[A]

Yeah.

[B]

But the scientists were not fully understanding how this is really happening. All the biochemical, biophysical properties, even though we were understanding, we were not able to predict the structure just by looking at the sequence. That was requiring huge computational power. That needs like extreme amount of power and time. But with the AI technologies that is that the Nobel Prize came to the alpha fold, where the linear structure can be given and what it will look like in the folded state could be predicted.

[A]

And help me understand why this problem matters. Is it easy for us to identify the linear state, but it's not easier? Like, why is this a hard. Why is this a problem worth solving?

[B]

The implications of proteins is in their functions. And the functions, as you said, is correlated. From their folded structure, we are able to get the sequence. And the sequence is coming from DNA. I think everyone is familiar with the DNA. DNA has a sequence. And there has been breakthroughs 20 years ago or that we were able to sequence the DNA. So we are so easily able to learn what the sequence is, but we cannot easily learn what is the folded structure.

[A]

But without that, we don't know the

[B]

function most of the time. Yes. Normally until this discoveries, it was taking a PhD student lifetime to discover the structure protein, literally. And if you discover true, then you're like, amazing.

[A]

Yeah.

[B]

It requires real experimental methods. There are different methods to do so. That is a huge amount of work. And to see how it looks like in another sequence, then you have to start over again. And that is important because as said, the implications are on the function. If the sequence changes, structure changes, function changes.

[A]

Yeah, got it. Okay. And then enzymes are a subset of proteins. Correct?

[B]

Correct. Enzymes are biological catalysts. So the proteins that has the function of catalyzing a reaction are called enzymes.

[A]

Yeah, got it. Okay. So I know like in the Nobel Prize, that group is using AI to predict structures from the linear sequence.

[B]

Correct.

[A]

And that's helping us understand function.

[B]

Correct.

[A]

Are there clear rules like you talked about? It used to take a PhD their lifetime to find one structure. Are there like, is this just a puzzle? And like we know the rule, are the rules chemical? And we know what structures are possible. And it's just like a brute force problem of finding the possibilities given the sequence. Tell me a little bit about why this is so hard.

[B]

Indeed, it was exactly like this. So we understand the biophysical forces and what it can do. But the co. The number of combinations and probabilities were so high to compute. It was a computation problem. And that is being sold by the neural networks and everything behind that AI tools that I admit that piece, it is happening at the software world. I don't have a good understanding, but what I know is the scientific rules that they base their work.

[A]

So the reason why I asked are there clear rules? Is because I can see very clearly how if you have a large neural net or in this case a large language model and it has clear rules of success and you can give it a sequence, we can use this same sort of predictive technology to just churn through. And if it has clear rules to test against, it knows exactly when it's succeeded.

[B]

Exactly.

[A]

Okay, so let's get into what you're doing with enzymes and recycling. So it sounds. I'm going to guess the goal is we want to find enzymes that have certain properties that allow us to break these polymers down into monomers with no waste. It sounds like there's two other variables. You're trying to keep temperature low because that's what reduces waste. You mentioned pressure. So I'm imagining there's a constraint around just the environment to reduce waste. So tell me a little bit about how AI is playing a role in this process.

[B]

AI is our tool to initially design new and better enzymes. Okay, so initially we use AI for the design of our catalyst.

[A]

Yeah.

[B]

And the vision we have is to take this to even to the more systems level where enzymes play. Enzymes are a player within a process. So understanding this different process components with the help of AI is where we want to take it. But right now we are using it to design our enzymes and maybe to give you a sense of why it matters. And what does that mean? I told you I did a lot of enzyme engineering in my past and I did it in old school. It is now old school. The traditional methods, which is mainly a trial and error, you just test things in the lab, you come up with an idea, you design it and you have to physically build this enzyme and go and test everything, which is time taking and also a lot of resource requiring process. Now we tackle most of this problem computationally using the machine learning tools and design new enzymes that the nature hasn't seen before. Nature didn't evolve to that direction. But machine learning with the algorithms can design this new enzymes and can present Them to us. And we can test only a subset, a small number in our lab, which accelerates our timelines and also decreases the resources that we use.

[A]

Yeah. So this is why I wanted to understand a little bit more of the science, because I feel like it gives a target for what the agent's trying to do. If I remember right, are enzymes amino acid chains.

[B]

Correct.

[A]

Okay. So amino acids are like our building block, our protein building block does the. And it's a sequence. So if we're familiar with DNA, it's the same kind of idea. Sequence of amino acids. I'm assuming the sequence is also. Same idea of the sequence influences the shape, which influences the function.

[B]

Correct.

[A]

You're looking for an enzyme that does a specific task. Maybe it's like we've got a polymer. We need an enzyme that can slice right here between this monomer and this monomer.

[B]

This is exactly what we do. But we also take it to the next level. Yes. That specific function it has to do. But it has to do it under certain conditions. Like maybe it has to be much more stable so that it survives longer. It has to do it. It can recognize certain structures, like materials have secondary structures, meaning that polymer itself is great, but it is not alone. How polymers interact with each other and form a secondary structure matters. Meaning that polymers can be very tightly packed or they can be more loose in their structure. That has an impact. And understanding this and designing an enzyme that can tackle this different polymer secondary structures are all within our framework of designing a new enzyme.

[A]

Yeah. And I can imagine you might need different enzymes for different types of plastics.

[B]

Correct?

[A]

Correct. Okay. So you talked about Arzu, which I can imagine was a very tedious process of. As a biologist, you have an understanding of amino acids. You have an understanding of how different chains, which then have different shapes, which then have different functions. You can have a fear, like a hypothesis, that, like, oh, if we chain this with this, it might have this shape, and it might do this thing in the polymer or in the polymer structure. And you have to literally, like, assemble those amino acids, run your experiment, and see did it do what you thought it did?

[B]

Yeah.

[A]

Okay. So from an AI standpoint, like, what I'm curious about, like, amino acid chains, like, we're still just talking about, like, chemical symbols. And so I can see this as. It's just a language. An LLM can learn that language. We're asking LLMs to predict tokens. I suspect we can ask it to predict the rest of the amino acid chain. Like, this part all translates really well in My brain, I think what, when I think about it from like a training standpoint or even like a reinforcement learning standpoint, are we just using like chemistry and biology, like the rules of chemistry and biology to evaluate the enzyme that's being designed and the LLM can look at? Okay, I designed, I produced an enzyme. But I like you're using the AI to filter what you test in the lab. What's the success criteria for the LLM? Is a lot of your work having to codify from a context standpoint what a successful enzyme is?

[C]

Before specifically answering that question, let's think about what kind of tools we are using. LLM, which is reflection of LLM in the protein world, is plm Protein language models. Okay, so there are protein language models, foundational models. We use those. That's one tool, foundational tool. We also use other architectures for doing other solving other parts of the problem space. Okay, I may go to the problem space one more time. The solving problem is one, meaning I have the sequence what is going to fold to. That's one problem space. Okay, once you know the fold, it's not enough to say this shape is going to have this activity, this function. There is still another problem space, secondary problem space. Okay, I know it's going to fall to this shape, but is it going to really do the function? I say required shape, but it's not enough to say it's going to that function. So there's a secondary second problem there. And the third problem is not problem, but the third module that we have is. Okay, I put everything together, it's all computational. I still want to limit my laboratory testing and cost. So we have a predictive model. Did we do the right thing? Can I Predict? So LLMs or protein language models comes into play at some aspects of this problem space. But not only tool that we use. Let me talk plms a little bit. Protein language models, you can imagine the more data to train the model is better. Therefore, protein language models created by using all the structural work that they have done traditionally, like our scientists done in the past, then they went beyond that and then they started using synthetically structured data. Long story short, the protein language models are trained with waste but so much generic protein data. In our case, you can imagine we are basically trying to develop an application specific application on top of these foundational models. What is my application? My application is specifically designing a protein enzyme for plastic polymer degradation. Okay, so you have these foundational models, but I need to come up with a abstract layer on top of it and maybe Multiple of them to create this application. I know I indirectly answered your question, but.

[A]

No, I love it. I love it. You did it better. You very clearly described what I was naively trying to guess. So it sounds like you have a protein language model that's being trained on what we know about proteins today. So it has some concept of sequences and what those sequences do that helps it generate a new sequence. Then there's this second model that you're using that takes that sequence and says, this is what we think the shape is. And then there's a third step which is given this shape, this is what we think the functions might be. Then there's a last step of if we did this in the lab, what do we think would happen?

[C]

Yes, slightly more complicated than that. In the first step, generation can be also multiple steps. Meaning, okay, am I in the directionally right direction in terms of a protein being a stable protein using maybe protein language models, but I also know a subset of those proteins are close to my enzyme, like K, for example, like a specific enzyme family.

[A]

Okay.

[C]

That I can also use either directly protein language models or just the embeddings of it to tell me the direction that I want to go in the generation can be also multiple steps using a foundational protein language model. I may use only after that the embeddings of it to. As an input to my own generation model. After that to come to the generation. Because again, when we are looking at it's not just folding, it needs to be like Arju said, we want thermostability, we want to be folded in a certain shape, et cetera. So within the generation, we can also not can. We are also using multiple layers and multiple models.

[A]

Okay. And I love that you framed this from like a foundational model standpoint. A lot of these individual pieces you're not building, they exist in the scientific community. And these you're leveraging. Just like an app developer is using OpenAI's AI API, you're using these foundational models and you're building the application layer on top. And your application is help me find enzymes that will do this specific job.

[C]

Exactly right. And even though we are calling it application, like you mentioned earlier, it has multiple pieces, Right? Like it's the generation first and then activity function again and then maybe there's another application. And then the third application is prediction, et cetera, et cetera.

[A]

Okay, so let's walk through this. In an application, I orchestrate LLM, not just LLM calls. If I'm building a rag step, I'm first get it grabbing and embedding. And then I'm storing it. And then I have even before that, I'm chunking it. Like I've got all these steps. And then maybe in a LLM call, the LLM is retrieving from that. It's looking at what it gets back. Like maybe let's talk through your whole sequence. Like I'm not sure I have a clear handle on what your input is. And then what are the steps that are. I'm getting a clearer idea of what the steps that are happening. And then I'm also really curious about this from a success criteria guardrail. Like how do you know it's doing the right thing? And this is particularly interesting because it sounds like the goal of this whole AI workflow is help us make better bets about what to test in the lab. And so I can imagine mistakes cost you time and money. So maybe walk me through what does that look like

[C]

going backwards. You said the right thing. Actually, we were just talking about it in the sense that our AI platform's customer is our lab and that our lab's customer is end to end process. So far we focus on delivering what our initial customer, our lab, requires. The lab wants to test minimum amount of enzymes and it wants to guess the highest performance metrics that we identified and can be multiple metrics, by the way. So we had done that actually in our platform version one that we call it. Let me describe that platform version one and then I will tell you the roadmap because perfect. Really fast. In the first version, again, we didn't only use LLMs, but let's call it intelligence or generation, et cetera piece. Right. There was a human orchestrating these pieces. Meaning are. So did the literature database collection. Okay. What do we know from the literature research, the nature shot, etc. So we did that within a human.

[A]

And then let me just. I want to make sure as you go through this that I understand each piece. I imagine that research is like, what are enzymes that have certain properties? Where are we getting insight from? What's you talked about enzyme families. If I'm thinking about this from an embedding space, what part of the room do we want to be in exactly?

[C]

From multiple different angles, specific to our function, specific to our polymer, specific to our family, the performance metrics they identified or what has been done in the literature. So that was step number one.

[A]

Okay.

[C]

And then again, I'm the orchestrator. There's a human orchestrator.

[A]

Yeah.

[C]

And then we have these other modules that that computationally we built including very simplistic again like three pieces maybe generation, validation and prediction. Maybe let's say. And then we built those, but every, each one of them I orchestrated that workflow.

[A]

Yeah.

[C]

And then output went into the lab and then we tested it. And then there is now private data that we have a benefit of that we fed into this floor.

[A]

Yep.

[C]

But more interesting thing with the latest agentic workflows and the power of these intelligence modules coming up now we are putting all these together without hormoning the loop maybe except the wet lab. And in the future we will also remove the wet lab too. But let's not get there right now. But currently our platform can generate this research that ARZU did in the past. Let's assume we're going to tackle another plastic target or we want to improve a different key performance for the enzyme we have on hand. So now we are putting all these pieces together with an orchestrator. Now also an intelligence, maybe LLM model, et cetera or AI scientists that can do all this by itself. And then we're going to be only in the loop for the wet lab.

[A]

Okay.

[C]

And at the input perspective, depending on our customer we may have different inputs. What I mean by that is Arjun mentioned that for pet, for polyester, which is our first plastic type. Nature showed us like eight years ago there was a bacteria found, et cetera. So that was our input. Meaning. Look, we know this enzyme has some activity, some characteristics that we want to improve on. It's not enough to use it at the industrial scale. So we know what we want to improve on. We want to improve thermal stability, we want to improve the yield or pace of the enzyme, etc. So those were our inputs. But you can imagine what we will do next is our input will be, look, we want to be able to break this polymer. I'm gonna break this bond now. Design me an enzyme that is gonna do that. Chemical catalytic activity, for example. That's also the direction.

[A]

Okay, so you're starting with here's a problem we're trying to solve. We need to break this polymer down. Here's the insight we had. We found an enzyme that's close, but maybe it needs to be modified in this direction. Go explore.

[C]

Yes, that is one way to do it. Then again it's going to get more and more powerful that we can define. For example, again I'm not coming from biology. The current understanding or current statistical is we cannot break or it's really hard to break carbon to carbon bond, for example. Enzymes can do that or very difficult to do is it really computationally we can define this problem and maybe our new artificial intelligence and AI scientists will be able to find a never existing enzyme with this flow. So we can start with this problem statement that we want to break new types of bonds, for example.

[A]

Yeah. Okay.

[B]

And then.

[A]

Okay, so you're starting with sort of a problem statement. You're putting it into this system. The system is generating sequences, predicting structure, making a prediction based on that structure. Tell me a little bit about guardrails. What's. How do you know that you're not going off track on each of those steps?

[C]

That's a good question. There are multiple pieces. One, we can define guardrails as an input too.

[A]

Okay.

[C]

Depending on, in this case, either research or Arzu's understanding of the enzyme. So it can be an initial guardrail at the beginning, and then for each step we can also have guardrails like again, based on her experience or what we know about folding, for example, we can say, you know what, even though folding is okay, we are looking for a certain shape. So every step that we defined based on what we know or what we expect, we can also define guardrails or we do define guardrails.

[A]

So based on the problem statement, you can add like criteria for the folding step. You can say it has to be a shape in this category or it has to meet this criteria.

[C]

Yes. And yes. And we can also do the guardrail to the output of folding state.

[A]

Okay.

[C]

Because since we can work with millions of sequences, thousands of them to start with, which was not possible in the traditional method. Yeah, Every step we can guide the input towards the guardrail and have the guardrails to the output of each module

[A]

and make sure it really is meeting the criteria that was specified as part of the problem.

[C]

Exactly.

[A]

Yes. And then what about the prediction piece? You're trying to predict what's going to happen in the lab? What's happening there?

[C]

That is the most exciting piece to me, but also the most challenging. It's a very known problem in AI. Not enough data. In our case, we are talking about hundreds, right? Like not thousands, not millions. But even though it's a low amount of data, since it's very specific, again, specific to our application, which means, okay, I want this polymer to degradation. So over the last year we collected hundreds of data points from your own wet lab experiment. From our own wet lab experiments. Exactly. Now I am using those in my prediction model for validation and testing and training, et cetera. So.

[A]

So the model might predict an enzyme that like on paper Looks good. And then you test it in the lab and you maybe identify something that makes it not work. You can feedback that into the model. And I can imagine you don't need that many data points before it can start to see patterns and see, oh, an enzyme with this property is disqualified for whatever reason. And so that every run through the whole loop that ends in a lab experiment, you're adding to this knowledge base specifically for the types of problems you're trying to solve.

[C]

True. It depends on the characteristic that we are trying to predict.

[A]

Okay.

[C]

So if the characteristic, let's say a catalytic activity is a basic one, there can be adjacent enzyme families that we can use the data or the model already learned. So for our niche case, maybe hundreds of them is good enough to predict the catalytic activity in a certain direction. But Arzu was talking about, we are taking it to the next step, which means we are not just trying to improve the enzyme performance, but also enzyme performance related to the process that we are using. That world is not shown in the training data at all. Therefore, we may need to produce a little bit more data points to make the prediction for that characteristic to be more powerful.

[A]

Yeah. One of the things I was thinking when Arzu, when you were originally describing like this research, I think you said it came out of Japan, but people weren't using it in industry. I know for a lot of, like lab science, there's a lot of work that then makes it viable in industry. It's not, oh, we proved the idea and here we are, it's successful. So I imagine that's a lot of what's happening you're hoping to do through this loop is it's not enough that it works. It has to work viably.

[B]

Correct. Correct. So the. Under the conditions we are working, the industrial conditions.

[A]

Yeah.

[B]

And also specifics around it, as said, the way the plastic is presented to the enzymes is an important factor. All these things taken into consideration for the design is taking it to the next step. So doing it in the lab is very different than doing it for industry. These are like one of them is very well defined and also very restricted conditions and parameters. The other side, the industry side is there is definitely still some definitions of specs, but has to work under much wider and different conditions.

[A]

Yeah.

[C]

On top of that, we are working with a commodity here, meaning our product is plastic. Yeah. If you go to Asia, they manufacture millions of pounds or hundreds of thousands of kilograms every year in a very cheap way from, unfortunately, oil. So not only it needs to work at the industrial level, it also needs to work at a business feasible way.

[A]

Yeah.

[C]

And it's not enough to say enzyme is doing its work and to end process needs to be cheaper too. Sustainability is great, but if we can, if we ask for 10 cents more, nobody will pay for 10 cents extra for a water bottle, for example, for sustainability. So that's why we are using AI too. We identify these key parameters, not just enzyme performance, but how it works with the process that is going to make the end customer happy, meaning cheaper.

[A]

Yeah.

[C]

And define use the AI to predict or to generate sequences or process parameters according to that end goal.

[A]

So what's fun for me about this podcast is I get to talk to teams across a wide variety of types of products and industries and a lot of themes emerge. A big one is context management. Like what are you giving the model to be good at what it's supposed to do. Another one is around. I think as part of context management, people build these like really complex, almost like data pyramids where they have like raw data and then they synthesize it at a level and then they synthesize it at another level. There's a lot around orchestration and like how much autonomy do you give the agent versus how much do you keep it on guardrails? And I'm just curious, like when you read about what other people are doing with agentic solutions, does it map to what you're doing? What are some of those equivalents? Or do you feel like this is really just a whole separate offshoot of what people are doing?

[C]

It is definitely not an offshoot. Actually the reason I was bought into this vision is, wow, when we can use these new tools coming in into this biology or computational biology, I think there's going to be a big disruption and Nobel Prize. What's validating that? So there's definitely a lot of overlap. Not maybe everything, but for example, orchestration is really big for us right now. Yeah, that's why I was describing our previous model that human me or Arzu was orchestrator. Now we are replacing that with maybe AI, scientist, model, etc. Etc. So yeah, estimation is huge for us because it is improving not only the final product, but also the development cycle for us. Because AI is a development platform for us. Yeah, we can make it more, faster or less. Computational is also beneficial. But also of course the end product will be better than what two of us can do. So orchestration is really key. And we talked about multiple pieces, right? Like we said data collection or literature, data accumulation or knowledge base. We talked about multiple different Problem spaces, all of them think about as different agents. But we need an orchestrator and then eventually the wet lab, maybe that's the human in the loop piece for us, but that also needs to come back into the orchestrator at one point and continue where it left off.

[A]

I also could imagine even your problem statements and how you're constructing them and the constraints that you define. And each piece of your pipeline or orchestration might need different context about the problem space. And that there'd be a lot of parallels with just what to give the models, when and how to move that through the whole system and models.

[C]

And again, since we are moving towards orchestration of multiple agents handling these different pieces, the orchestrator needs to understand the full context, but also provide only the necessary or important context to each agent. For example, okay, I have a module, I have an agent who only takes care of the folding, for example. Then it doesn't need to know anything else, but it really needs to know what are my guardrails or direction for the folding.

[A]

You described this initially almost as a deterministic pipeline, but I could imagine if the folding agent like hits a constraint, it might have to go back to the generation piece. And even in the prediction, you violated a constraint, you got to go back to, maybe even back to the sequence. Is that now that you're moving to this orchestrator agent, is that kind of what it's. Is it coordinating between the steps?

[C]

A hundred percent. But I also want to coordinate something even though it looks like deterministic. Nothing in this flow is deterministic. Yeah, you are folding. It's still a statistical model.

[B]

Right.

[C]

Compared to what humans used to be able to do in the past, it is pretty deterministic.

[A]

But I, I meant, Sorry, I meant deterministic in terms of first the sequence is generated, then the F molding prediction happens. Not that within this step, it's deterministic.

[C]

That's right. That's right.

[A]

Like a lot of teams have gone through this with just even like web features. They start with a pipeline. Like each step might be an LLM call, but it's a. It's not agentic. Right. It's just a pipeline. And then as the models get better, teams are moving more towards, let's actually just have an agent and each step of the pipeline is a tool. And now the agent can decide, am I ready to go to the next step? The next step broke. I got to go back to step one and you get a little bit more of like non determinism at the flow level.

[C]

Exactly. And further Than that. For example, we are simplifying some of these modules. Let's say we are calling about alpha for folding. But for folding problem. I can also have other tools is a signal. So agent. My folding agent can use AlphaFold but also can get more other tools to use get more signal. So that is also an autonomy to the folding agent to do with the goal and maybe multiple tools and then get back to orchestrator with the potential results or findings.

[A]

Okay, so we get all the way through your process and outcomes. I'm assuming the output is. Here's a compelling enzyme experiment. Give me a sense of what has been the impact of having this. Are you able to run more wet lab experiments? Is the goal to run fewer and have a higher hit rate? What's been the impact of having an AI scientist on your team?

[B]

The impact is again on definitely the timelines and also the diversity and the design space. We can get it. That is I will make the. I will only able to explain it with a comparison to the old school way of doing it. In old school way generating diversity was a stepwise thing. When you engineer enzymes, you identified positions and the amino acids changes that make an impact and then you go and combine them and. But every time identifying takes a lot of time. And then combining that takes another amount of time and resources. Now we bypass all of this and get different diversity of design in our hands to go and test. And that is like changing the paradigm in a way and also the changing the timelines. Because the amount of time and resource it takes for MERT is completely different than what it takes if you do it. The old school, which is literally manual human predictions and suggestions and then go and test them in the lab. The other alternative is doing it really more like a gunshot that just without putting any rationality just test many different versions without knowing where it will end up and. But then it cost a lot of money because you need. You need to test in physical world. So definitely the timelines and the diversity. The what? I mean diversity is the space we end up in. It is not like an incremental thing. We end up being in a completely different design.

[A]

I can imagine in the manual process a lot of a scientist like hypotheses about different enzymes is really influenced by the literature. They followed their past experience and that the diversity is probably pretty low just because of like human limits on what we can know. Right?

[B]

Yeah. The way to search for diversity was actually going to the nature. That is how we created diversity. We knew that we are limited in the number of Sequences and diversity in our hands in the lab. But let's go to the nature. Nature has evolved many different organisms, go and look what they have generated and copy from them.

[A]

That is interesting.

[B]

Experimentally we did to generate diversity. Now we have a tool to generate diversity in a much more quicker manner.

[A]

And I can imagine like all that knowledge of what we know from nature is what's going into that protein.

[B]

Totally. That is what makes it possible to begin with, but allows us to go to other places. And generally I like to present it this way that the design space is huge. Think of it, this whole universe and nature has evolved to that much of that design space. Now we are able to explore all the other pieces that the nature hasn't evolved into. And that is, for me, it's mind blowing.

[A]

I would say that's a really great point. Just thinking about, at the human scale, thinking about taking inspiration from nature, that feels so big. But what you just said, I want to just highlight that this space is even bigger and the L and the models have the capacity to explore this even bigger space. So like the space that felt overwhelming to us was a pin. That's very cool. That was very awe inspiring.

[B]

Yeah, because as said, our reference point is always nature in biology. Like you go to nature, you start from there, come to lab and start engineering in a rational way. That is what we did. And if you want to generate more diversity, you were going back to nature and explore different environments to create that diversity. But you, as said, you are still limited within a space there and AI is bringing something else. And I'm really curious to see where things will go and end up. We are, I think, at the very early stages of all these developments.

[A]

It's fascinating that this nature has benefited from millions, maybe billions of years of evolution and yet it's still like a tiny percentage of the total search space. And like I'm having this moment of like when you read, when you learn about my cosmos and you get a sense of just how big it is and like you start to struggle to wrap your brain around it. It's a little bit of the same thing we think about the diversity of living organisms. And that already is huge. And now you're telling me, but it's just a tiny piece of what we could be exploring from an enzyme standpoint. That's amazing. I love that. Mert, you looked like you wanted to jump in earlier.

[C]

I was going to say build on that. I'm curious because you said you also talked to other product, for example, teams, et cetera. I'm sure the LLM hallucination is a problem.

[A]

Yes.

[C]

In some cases it's not for us. And I was just thinking, for example, in some applications or some modules, I want it to hallucinate.

[A]

Yeah.

[C]

Because if I provide the existing research as a context, like she said, then I am also limiting what you guys just talked about. Right. We want to explore that design space. So in some cases the hallucination is a. You want a high temperature that I want to use. Yeah.

[A]

You want the creativity mode.

[C]

Exactly.

[A]

Yeah. This is great. Okay, tell me a little bit about what's next. It sounds like you've got a pretty good system in place for predicting enzyme designs. I imagine you could iterate on that infinitely. But you alluded to. There's more here.

[C]

Like I said, the next one is instead of thinking our customer, a computational biology customer, AI customer, the enzyme design making to the next one, which is the ultimate goal, having an end to end process that is feasible business wise. So currently our startup, we are scaling up this process to 5,000 ton a year, like a demo plant in California. So we are developing this orchestrator model, agentic model, not only enzyme design, but we are introducing like a process agent, for example, which understands our process and provides inputs or develops maybe new process parameters. So that's the next step which is introducing this process angle to it, but also make everything more autonomous and loop and ideally not right away, but we also want to get rid of the wet lab testing. At one point. If my predictions are strong enough. Right. Then I will reduce that. Our predictions are strong enough, we will reduce that significantly. So.

[A]

So if I remember right from your application, you're partnering with a bottle lab in California.

[C]

That's correct. So that's how we gonna scale up our process in California to a demo plant, plants.

[A]

Okay. And the goal there from a process standpoint is it's great that we have this like prediction to get to an enzyme, but there's the back half of the problem of we've got our reactor, we've got our enzymes working. How does this fit in the bottling plants ecosystem?

[C]

Exactly. Even within our process, after we break down those monomers, we need to separate them. So that's another process. When we are doing the separation, we are realizing there are some effects of enzyme in the downstream separation process. So there are new potential process or enzyme parameters that we can focus our AI to target, which are not a typical enzyme design, for example, which is more specific for our process.

[A]

Okay. I could also imagine if you get to a point where you can really scale this. You almost could do like custom enzyme design for a customer. Right. Like a customer's product is specific, designed in a specific way. You're designing an enzyme specifically for that product. And then this process agent can also learn about specifically how that product is recycled. And you're almost getting to bespoke enzyme design and process integration because the AI is driving all of it. Is that part of your vision?

[C]

That's absolutely correct. Because we talked about the foundational level, our application, and that depends on the abstraction level, like how much we abstract or not. Yeah, our next step is not any enzyme, but our next step is different enzymes for different types of plastic. You asked that question and Arzu mentioned the biggest problem in the plastic recycling is the material is mixed. You will have a few different enzymes for a few different plastic types. Therefore, when those plastic types are together, you don't need to separate them. Our enzyme gland will tackle all of them. So we're going to take our platform to go to developing new enzymes for new different plastic types. And after that, potentially there can be some different enzyme applications. But yeah, there are enough, unfortunately enough different types of plastic which will keep us busy for a while.

[A]

Yeah, I love this. Okay, I want to end with a question for Arzu. You showed us a blueberry clamshell. In my town, clamshells are not recyclable. Are you going to solve, are you

[B]

going to solve that one?

[A]

What's that?

[B]

You touched a very important and a good point. I wouldn't like to get into it because I thought this is very domain specific problem. So this material I was telling you about the secondary structures and the secondary properties of materials, this material is different than the water bottle material. Even though they are the same chemistry. Because this has already lower molecular chain lengths. That is why they are not recyclable. Not that they are not recyclable. They are not favored as a recyclable material because to begin with the chain lines are short and the process generates shorter chain lines so that the quality is even lower. So this is just simply not favored. But the bottles are have, like to start with, have longer chain lengths and you can break them down a couple of times until you get really bad material. So for us, this material and the water bottle, they are all the same. Actually the polyester T shirts are all the same. It is, we break them down all the way to the beginning. So like just an one of the

[A]

biggest curiosities I've had for a long time. Every time I throw away a clamshell. I wonder why can't I recycle this? So thank you for that. Yeah, this has been amazing. I've really enjoyed learning about your story and learning about your company. You give me so much optimism for maybe we can actually solve this horrendous plastic world we've created. So I am going to be cheering for you and I really appreciate you taking the time to share your story.

[C]

Thank you.

[B]

Thank you so much, Teresa. We appreciate too.

[A]

If you enjoyed this conversation, please subscribe in your favorite podcast app and give us a rating as it helps others find the show. Thanks. I appreciate it.