Brandy Whalen (00:02.327)

It is a pleasure to welcome Reza Farahani to the show today. Reza is a pioneering force at Katalyze AI where he blends biotechnology and artificial intelligence to reshape the future of manufacturing. With deep expertise in approximation algorithms and data analysis, honed through advisory roles at Avery and dynamic collaborations with partners like Walla Leeds, he brings a unique

market design perspective that's driving transformation change in the biotech sector. Reza, we are thrilled to chat with you today. Thank you so much for joining us.

Reza Farahani (00:40.056)

Thank you, Brandy. Thanks for the intro and thanks for having me.

Brandy Whalen (00:43.297)

Yeah, absolutely. Well, let's jump right in here. Can you just tell us about your background and the journey that led you to found found Katalyze AI?

Reza Farahani (00:52.878)

Yeah, for sure. I'll come from an engineering background. I'm a builder more than anything. I started my journey as an engineer data scientist around 2015, 2016. And after a while, doing data science, ML AI work, at a time you needed to write this neural net stuff from scratch, I...

move and join to the Boston Consulting Group, BCG, as an AI expert in Toronto. I was working with them for a couple of years. A lot of the stuff under the moon I did like, know, enterprise AI building solutions for financial sectors, for manufacturing, like CPG, consumer product goods.

always building those like in that building mode even when I was like consulting. Left during COVID, started my journey as an entrepreneur. I had my first startup called Work From Home. That was a response to the remote work happened after COVID. We do take the company from zero to half a million user in only 18 months, raised a couple of million dollars and had an exit.

after just 18 months and after that it started Katalyzed. That's kind of like a quick journey of me always in the builder mode from the starting point, you know, up to now. Yeah.

Brandy Whalen (02:21.975)

Yeah, well, and it's a little bit of a pivot, right, into this world. So what was the inspiration behind launching Katalyze and how has that vision evolved?

Reza Farahani (02:23.822)

Cough

Reza Farahani (02:32.62)

Yeah, so, Cattle is an interesting idea and I, you know, being biased as a founder here, but the whole space of bio-manufacturing is very interesting to me. The reason behind it is like, when we started bio-manufacturing that get back to probably like, know, during the world war, when they wanted to start producing penicillin, we have used that format, what we call it the precision fermentation.

What does that mean is like we are using a live organism to produce specific molecules. And inherently as in like someone who's like interested in the engineering and chemistry, that's very, very interesting. So we are, your unit of manufacturing is a live organism, is a cellular, is a cell. But what are the use cases here that was even more interesting?

So when we start using it, we mostly focus on the therapeutic, you know, drugs to use by manufacturing. But as we are moving forward, we are seeing more and more use cases coming out. That is not even a therapeutic, which again, like therapeutic is really important, but you are seeing. have product companies like Bon Vivant or Zero Milk Cow, the companies that making milk through bio manufacturing. So animal product without having an animal. So that would be much more sustainable.

Brandy Whalen (03:52.153)

interesting.

Reza Farahani (03:57.998)

much more costly, the quality gonna be good. We are seeing companies making like a gelator, know, the CEO, I have like a few conversations with there. They're making gelator, like you know, the gelatin is being used a lot of cosmetic gonna be like it's stuff paid coming from by manufacturing. And in general, we are getting good precision medicine like you are having it like a drug that is.

Created only for you because you know, that's just designed for you and making that drug needs like sophisticated bio manufacturing So we are saying that's like so much use cases in this space. We are if I want to like make an analogy is like we are in a 80s of computer that we use by use by a manufacturing today for stuff that are very very expensive like Kachruda oncology drugs by Merck like the spring my manufacture But in future we are gonna see

Brandy Whalen (04:40.248)

Yeah.

Reza Farahani (04:52.512)

when the price drops, as you're seeing in that, but we're more and more prevalent in the consumer sector, seeing stuff you're using day to day being by manufacturing. So that was a big, big idea. was like, I'm really interested in that idea. And I think the price of that by manufacturing going to drop drastically. And we will Katalyze what it'd be like to facilitator of that. We're going to like help the price drop by 10 X in the next five years.

Brandy Whalen (05:18.849)

And so do you see that as kind of being the number one problem that Katalyze is solving for is to decrease that the price in bio manufacturing?

Reza Farahani (05:30.67)

I wouldn't say like decreasing the price in a sense of making more accessible. So when you come up with anything like, hey, is it possible to make paper by buying a refraction and you figured it out or you make a drug, you know, like, I found a drug for a specific disease. Even if it's a rare disease, like, and you want to like manufacture it right now, there is so much cost in the process development. And sometimes, you know, it's not about like making it cheaper. It's like that

Brandy Whalen (05:52.536)

Yeah.

Reza Farahani (05:57.358)

price decrease or that availability is between having a drug for a rare disease, a disease that like thousand people have because it's so expensive, so time consuming. It's not like, you know, financially viable to make it or having that drug for like that those thousand people. So it's not only the like a cost in a sense of making it cheaper. It's just that's a break, like a tipping point between having something and not having something.

Brandy Whalen (06:25.303)

Yeah, so it's not even necessarily, I hear you, 100%. So it's really just creating an available market, right, for these things that otherwise it doesn't make sense to make. mean, you're talking about any sort of rare disease, personalization of medicine. Right now, the hurdle being there's not enough there for it to be impactful, right? They just can't, it doesn't make any sense.

Reza Farahani (06:52.276)

Exactly. like use cases are so prevalent so much. Yeah, we are talking to companies that are making like right now vaccines that are mRNA vaccines for Actec to make it like, you know, resistance to You know the diseases that they get so that's gonna be like it's very futuristic like no Pesticide, you know getting used there and you still get get your like, you know

can get your product without any issues, without any disease. So there are many use cases exist that even today we are not like, we don't fathom that, we don't understand it. But if you make that technology available, like when you start making computer, no one think about thinking about the social media as a technology. So to us, this is like a very fundamental technology that can make production of complex molecules available.

Brandy Whalen (07:31.235)

Yeah.

Reza Farahani (07:48.48)

available and you know and what those complex molecules do it's up to the like we don't know what are those what are the use cases we only want to do may we the our goal is to be the infrastructure to make that possible for everyone who had a vision there

Brandy Whalen (08:08.311)

Yeah, absolutely. That makes complete sense. And I'm just kind of thinking here, Reza, just what would you say are some of the biggest challenges bio manufacturing faces today? And how is AI data, how is all of this kind of uniquely positioned to address them?

Reza Farahani (08:30.082)

Yeah, it would be really interesting to even look at the history of bio manufacturing. When they started by manufacturing penicillin, the amount of penicillin in the broth was less than gold in the ocean water. and the first person, when they, a fellow man figured out like how penicillin can be useful to cure, you know, this is as a...

They couldn't manufacture in Europe, they send it here and a few companies like the parent company of BMS, Pfizer and Merck were working on it. And this guy was coming, like they couldn't make it synthetically through like chemistry processes. So people thought he's crazy, that's little bit suggesting that we need to do bio manufacturing, like, know, getting gold out of ocean water. However,

Today we are 50,000 times better than that. Like the amount of penicillin that we are harvesting from the broth is 50,000 times more prevalent.

It's still, it's very expensive, very volatile. And the reason behind it, because you are living with, dealing with a live organism. It's a series of engineering, you know, issues that you are like facing and every day. Yeah. Even if you talk to the people who do like a not precision fermentation, biomass fermentation, like they can tell you, like if you are brewing your beer,

The pressure of the air can change your fermentation process. So delicate. So you are seeing with, when you're talking to this big bio manufacturing companies like, Hey, like our target is our theoretical target is like 400 grams per batch. We are getting foreground. The gap between what your theoretical limit is and what you're doing is like, like so much. And between that target, even that you put towards yourself, like that's one person target that you put for yourself.

Brandy Whalen (10:00.747)

Yeah, so delicate.

Reza Farahani (10:26.754)

you are seeing there is variation. You see a high level of variation. If you are putting the sugar as a source of energy for your microbial and that sugar is like the crystal size is a little bigger, maybe it doesn't solve as much as you want. And that can change your whole processes. So there is a lot of variation. There is a lot of sensitivity because you are dealing with the live organism. This is inherently a hard thing to figure out. And it's really difficult for a...

Human because the amount of data amount of variation is for the people who understand like data science the dimensionality of the problem is too high for a person to fathom it so the only way to deal with this one is a data-driven way and using like advanced algorithmics to monitor predict and recommend solution

Brandy Whalen (11:18.719)

So what is the reality? Because yeah, I see that just as being a massive advantage to be able to kind of push forward and allow for more production to happen in some of these variable environments. So what are some of the, what's happening currently? Like, what does that look like? you've got just, you've got a production yield expectation and you're just not meeting that. Like, what does this look like for companies?

Reza Farahani (11:47.948)

Yeah, I mean if you guys follow the news you can see like we are in the biggest shortage of drugs currently in the US I would say in the world. Yeah out of top 20 drugs 17 of them are by manufactured and like there are like half of them are in the shortage like if you make one new dose of like specific vaccines, there's a buyer on the other side, so What you're gonna do either you are gonna

Invest billions of dollars to make like open up new facilities. So that's very common right now It's like hey go and the companies cannot like afford it usually like there is so much subsidies coming in like taxpayers money's coming in you're like invest in a facility billions of dollars to just Open up a new facilities are big facilities like so big like these tanks are so big to take them out You need to tear down a wall to take them out and bring it back in very very costly

Brandy Whalen (12:30.04)

Yeah.

Reza Farahani (12:45.484)

And for optimizing your current process, it's usually throwing people at problem. It's you there, you hire more scientists, you hire consultant, all through edit problem, you just find the point solutions like, hey, maybe we need to change this filter today. Like I'd change it to another type of filter, maybe we get like a couple percent more on yield. And most of the time it pays for it, even if you spend millions of dollars on like studying at a filter.

and you change it and you get like a 3 % or 4 % lift, that pays for itself. However, like mostly the current situation is dealing with it, like throwing people at the problem. And honestly, like we have shortage of people who understand the topic. How many people understand the engineering science on a scale of biomanufacturing? Very, very little like number of people who can that capability. And you see these people sometimes work 80 hours a week to just, you know, make sure

Brandy Whalen (13:26.169)

Yeah.

Reza Farahani (13:41.592)

plant, a factory is keep and running. And this is a very noble job, very difficult to do, but that's a current situation that we are dealing with.

Brandy Whalen (13:51.469)

Yeah, I also think I could see this kind of coming into play and being hugely beneficial in the regulatory space too, right? Because in bio manufacturing, there's oftentimes very complex regulatory requirements. So how do you see the data, the capabilities kind of playing into compliance and especially in the quality assurance?

Reza Farahani (14:18.048)

Yeah, one interesting thing is like...

because of the regulatory, because of what we are seeing in the regulatory space. A lot of the efforts of these engineers, scientists go through making documents, keeping track of what's happening. So a lot of human effort, a lot of human hours go, like human hours are like very, very talented. It's scarce resources go through making these documentations for FDA, for like regulatory, for quality assurance, as what you said.

And we believe this is a job that can be automated a lot with the data. And we can be automated even more. And it's been done in a more accurate way. we can, the AI, know, the automation solution can do this job cheaper and better than a human and free up like very, very scarce resources for us. And we are seeing this one coming to the market. know.

I was just reading a news yesterday that no one or this for all Zempic like, know, the reports around annual quality report that need to be filed for FDA. Now they're using AI for that. And that process is fully being GMP validated. Like, you know, we are seeing this one is coming up and that's very, very good news for the people doing this one because that's something we need to do. Honestly, I would not advocate just reducing any regulatory for the manufacturing of

you know, vaccines, and therapeutic. So the other ways that using AI to make that easier for everyone look and, you know, reduce the cost.

Brandy Whalen (16:01.143)

Yeah, yeah. And as we as we look into the future, Reza, especially in this space, like what are you what are you seeing? What advancements are you are you most excited about? What what kind of transformations do you anticipate happening?

Reza Farahani (16:17.262)

Yeah, one thing, there are a few things. One of them is we are seeing more and more on the hardware side, you're tabletop fermentation, like or precision fermentation tool. That's what it is, like enables, you know, a lab and a startup with even like 10 to 20 people, a couple million dollars in there, like bank account, kind of start experimenting with the fermentation.

Because if they don't do it, they need to outsource it. That would be very costly. They lose a lot of their patent, their IP to the third party that is manufacturing it. So that's become more common. We are going to see a boom in, you know, I would call it like the indie bio people, like that are not the big pharma companies. That's very exciting. And the second thing for us is like, that's what we are doing. are going to, we are looking towards that democratizing.

Biomanufacturing like this the same situation with the tabletop What is like if you get that tabletop like what what's what's the recipe? Yeah Do you need to go higher like really like couple expensive? fermentation scientists with like years of experiment in making this drug to come like making going through this process to come and help you to do the process development that can take like 18 months hundreds of runs and millions of dollars or you have the table top fermentation

Brandy Whalen (17:13.945)

Yeah.

Reza Farahani (17:41.794)

You have the Katalyzed software. If you say like, hey, this is the target molecule that I want to make. This is my cell line. This is like the vector. This is the instruction. SPW recipe for you. So we are working on a zero-shot learning model, like very similar to that LLM model that you are seeing day to day that generates the, you know, paragraph for you, some wording for you. are generating the recipe from scratch based on the information we have.

on the target molecule and also your facility. So that these two are gonna be democratizing the bio manufacturing. And that's what I'm saying. Like we are gonna see the boom reducing the cost for people's like, for startups to get into this space without passing those hurdles of having a facility or hiring a third party.

Brandy Whalen (18:35.043)

Yeah, and I would say that that's usually what crumbles companies, right? It's just that it's so cost prohibitive. And when you're talking about like indie bio manufacturing companies, so startups, they're looking, that's where innovation happens. So you're allowing more of that to take place and eliminating some of those barriers to entry, which is incredible.

Reza Farahani (19:01.454)

Yeah, exactly. Like giving computer to people and they start coding and they're figuring stuff out just enabling these people to like think about like, hey, I make this. I think this is a very like good use case to make. Again, like I'm going to write a gelatin like without pig, but like how are we to commercialize it? How are we going to make it produce? If you can reduce that time and you see more of this company coming on, like, you know, there's a lot of animal product that going on.

that could be done so much cheaper, so much better. We are talking about silk, are talking about meat. We are talking about very, very interesting use cases that I hear every day because I'm in the community. like, this is super exciting. What provides you from taking that to the markets? The cost of figuring out how to make this one in a commercial level, that's the problem. If you figure that problem out, there is so much stuff going to get resolved and the rest of the diamond is just going to fall through for them.

Brandy Whalen (20:01.655)

Yeah, and I would imagine too, Reza, you've like strategic partnerships are really incredibly valuable to you, especially at this point and always, right? Are there some key partnerships and collaborations that have been kind of driving the innovation at Katalyze?

Reza Farahani (20:13.091)

Yeah.

Reza Farahani (20:21.368)

Yeah, definitely. We cannot do this without the collaboration. This is like a very, it's a big ambition and cannot be done alone. I just say, like we have a lot of collaboration with the big pharma companies. We work closely with Sanofi, one of the big pharma manufacturing based on France. We work closely with a few universities, both in Canada and in California.

Brandy Whalen (20:29.283)

Yeah.

Reza Farahani (20:50.644)

and work with even like, you know, companies that are making, helping us like to get like computation power. We like AWS, NVIDIA, like those are like very, very crucial resources for us, for our team to do our development better to do their, like, you know, getting access to more data. So data and computation are the two key factors for us. And like having those types of partnerships with like

Brandy Whalen (21:15.502)

Yeah.

Reza Farahani (21:19.054)

big provider of chips in the world and a big provider of the data in the world was so helpful so far for us.

Brandy Whalen (21:26.083)

Yeah, that's really incredible. And even just kind of thinking about you as a leader, your vision, as you started Katalyze, you're moving through, growing, launching additional products. What lessons have you learned as a founder and a leader in this space?

Reza Farahani (21:47.262)

And there's a lot of learning there, obviously. think like when you, I would say like, starting, like when I started the company, would say the first thing is just starting the company is very, very difficult. So a lot of you, I've been into like corporate environment. I've been to your academic environment. You see a lot of people talking about ideas.

Brandy Whalen (21:50.272)

I bet.

Reza Farahani (22:13.79)

and very few people have the courage to jump in and starting an idea, like working on it. So I would say that's number one. I like, I encourage, I wanna see more people doing this. I wanna see more people getting to the weeds and starting a company doing this stuff. And at least experience how does it work? Because being in the other side just gives you a different perspective on life, different perspective on like how business works. The second...

Brandy Whalen (22:17.805)

Yeah.

Reza Farahani (22:42.19)

Lesson I learned is like, you you take care of the people and the people your people take care of your company that's a big one like we try to After a certain, you know number of people that you pass You don't have a full grasp of like every piece of the company you do You just want to have people that do the job the best maybe like better than you and in this, you know, I like

in the sector that you're doing. I cannot do the both coding and also like all the sales and everything. So you want to make sure you find right people that do the job better than you and give them and train them like and you know, give them the freedom to do their job in the best way possible. So if you do put these pieces together, like you reach to a point like, okay, yeah, you can run a company and you can focus on a high value items that you think

Brandy Whalen (23:16.429)

Yeah.

Reza Farahani (23:34.764)

You can drive the vision of the company. So it gives you that freedom to do your job better.

Brandy Whalen (23:42.103)

Yeah, and kind of even just thinking about what type of advice would you give to folks in biomanufacturing? you know, it's just getting a kind of wrapping their arms around the capabilities and what they can start doing with the data, with the information that they already have. Like, how do you get them to kind of just take the next step, take the plunge?

Reza Farahani (24:08.674)

Yeah, I mean like the thing that I want to say is that just try experimenting with the stuff. And that could be even that you can invest if you have the capability in building a stuff, that's fine. But there are so many interesting tool out there that can help you to do your job better. Just have the courage to take the risk and try stuff out. And that's a big one because we are in this space that like

There is a lot of moving parts happening. And if you don't get caught up with the pace that the technology is moving forward, you are going to be the lagger and you are going to pay that off one way or another in the next couple of years. We are seeing how some of the players who adopted the new technologies, the new AI just moving so fast, so quickly.

Brandy Whalen (24:56.259)

Yeah.

Reza Farahani (25:07.246)

and how the ones that are not there like falling behind. It's very, very obvious that the speed of execution matters and AI is drastically improved the speed of execution. We are seeing, and this is even beyond bio, we are seeing more and more companies that are coming with a, you know, $100 million ARR, billion ARR, with such small teams.

that is built around AI. So if you are a heavy, you know...

organization with a lot of waste and not adapting AI, you're going to pay that off in future because we are seeing so many lean companies, natively built around AI doing the job that these big conglomerates are doing so much cheaper, so much faster.

Brandy Whalen (25:59.289)

Yeah, no, I think that's great. That's great advice. What's next? What's on the horizon for Katalyze? Are there any new products, initiatives, announcements that you'd like to talk about?

Reza Farahani (26:10.862)

Yeah, yeah, we have we are working actually like these days on a very exciting tool we call it like agentic root cause analysis I know the agents are getting very hyped these days, but the whole idea behind that is What I talked about around FDA and regulatory is is whenever an issue in a manufacturing facility happens You need to go and I started doing the root cause analysis on that

figuring out what was the issue, solving a problem. And this is gonna be the next big thing for us that we are working on. An agent that kind of reacts as your team of scientists and data scientists. So anything that happens in your facility and you have a question why that's happening, that agent can go to find the root causes of it. That agent is built on top of our in silico model of manufacturing.

So it can run different simulations over and over and over and see what are the possibilities that could cause this problem and start identifying them. On top of that statistical model, can take a look at, it can reason on the biophysics element. So for example, if you are getting a result that says the temperature of fermentation tank is rising, it's going to look at like the possible

Brandy Whalen (27:28.337)

wow.

Reza Farahani (27:37.878)

scenarios, running SiliconeMollet to verify it and come up with a reasoning and generate a report automatically for you out of the box. So we are going to launch that product in the next couple of months. We are very excited about it and we have really good partnerships on our side.

Brandy Whalen (27:55.789)

Yeah, that sounds incredible. That's amazing. You know, as we're wrapping up this conversation, where can people find you? Like, how can people take action? What's the best way for folks to get more information, to engage with you, to engage with Katalyze?

Reza Farahani (28:14.4)

Yeah, for sure. are, you can reach out to us at our website is Katalysai.com, Katalyse with K. We have our LinkedIn accounts, the same, Katalysai.com at LinkedIn and Twitter, same situation. We have, they can reach us to us or reach out to our team directly if they have any question. If you have a newsletter, we have a podcast, a lot of stuff going on there. I think

There's not much information in this space of bio manufacturing. They can found very, very interesting content there. If they're interested in a space, interested in this space, the intersection of bio and AI, we are talking to very interesting people. So yeah, check out the website, check out our LinkedIn, our Twitter, and they can reach out to us directly or via the forums in the website.

Brandy Whalen (29:05.569)

Amazing. It was a real pleasure to be able to speak with you, being the founder of Katalyze. And as many of the listeners know, this is a podcast that is brought to everyone by Katalyze. And we really focus on leaders in this space who are at this intersection of AI and bio manufacturing. So it feels like a real treat to be able to feature your voice today. So thank you so much, Reza.

Reza Farahani (29:35.406)

Thank you, Brandy. Take care.

Brandy Whalen (29:36.855)

All right. Yep. Thank you.