Podcasts

Podcast with Elica Kyoseva, Quantum Computing Scientist at Boehringer Ingelheim

29
December
,
2021

My guest today is Elica Kyoseva, Quantum Computing Scientist at Boehringer Ingelheim. Elica and I talk about quantum computing in the pharma industry, the composition of their quantum team, the degree of competition and collaboration between companies, and much more

Listen to additional podcasts here

THE FULL TRANSCRIPT IS BELOW

Yuval Boger (CMO, Classiq): Hello, Elica. And thanks for joining me today.

Elica Kyoseva (Boehringer Ingelheim): Thanks, Yuval. Thank you very much for having me at your very cool podcast.

Yuval: Who are you and what do you do?

Elica: My name is Elica Kyoseva and I'm a quantum computing scientist at a German pharmaceutical company, Boehringer Ingelheim. previously, I was in academia and I have been working on quantum computing on several continents, including Asia, North America, and currently The Middle East, and based in Israel.

Yuval: Working in pharmaceuticals, I think there are a lot of use cases that people talk about why it could be fantastic for their future, for pharmaceuticals. But is there something that can be done today with quantum computing or do you think pharmaceutical companies are just preparing for distant future?

Elica: it's an excellent question. And the truth is that the biggest part of our effort is really to prepare for the distant future, which in our eyes is of course fault-tolerant error-corrected quantum computers. So, the majority of our use cases and really the problems that we are looking at are facing and answering the question of, will it be possible to implement these algorithms on the fault-tolerant quantum computer? However, even very recently, two weeks ago, we actually released a paper with one of our partners, QCware. And it is about NISQ calculations, where we show that using NISQ quantum computers can bring improvement to calculating binding affinity. So, this is really a step forward that is showing already tremendous progress for the NISQ era.

Yuval: How large is the team? 

Elica: It is not a secret. We are five people, which is actually pretty large. I know quite a few companies that are already ramping up and hiring as well. So soon, this will not be a big news, but we were probably the biggest team in the industries, especially in pharma. And all of us have PhDs and we all come from the traditional way of going through the universities, getting technical training, and then we kind of diverge. So, we have a person that was a postdoc at Harvard. I was personally even at the venture capital fund before I joined BI. We have IBM folks again in the university of Bristol. So, we are a diverse team, but I would say everyone is quite highly, technically skilled.

Yuval: When you talk about diversity, is it also diversity in education? I mean, putting aside where the education was. So, for instance, you have computer scientists and pharmaceuticals, or everyone has a computer science or physics major, what is the sort of structural composition?

Elica: So, from my perspective is it diverse, but it is diverse only from the quantum computing stack. So, we do not have people that are really, let's say, classically trained quantum chemists. The focus of our training was quantum computing. But basically it's zooming in on different parts of the quantum computing stack, all the way from algorithms and implementations and applications, but also all the way down to the hardware. So, personally, my background was in quantum control and I was looking at quantum optical problems of how can we control the qubits in a robust way. So, I was really looking at error mitigation of the quantum hardware.

Yuval: Do You think that full stack experience is going to be necessary in the future? I mean, I think you're quite unique that you work with the hardware and then the software and so on, but wouldn't people want to get more abstraction and say “ well, yeah, I understand what a qubit is, but I don't really care what the pulses underneath look like”.

Elica: Ultimately, I think it would be important to do both and software and hardware,  to meet in the middle. So, it is nice to abstract our algorithms and really to focus on applications. However, this can become quite detached from reality very quickly. Even if you think about many of the NISQ works and the NISQ algorithms, they are still noiseless. So, for many of these results, actually there are no noise models that are implemented on the system. So, ultimately it is very important for people to keep a connection to reality, and especially to the lifetime or the qubit and exactly of the different noise processes that are in the system and design the algorithms with that in mind. Of course, this is for NISQ algorithms for the ones that we want to see in the near term. For fault-tolerant, I think, there you can still be quite abstract and focus on the algorithmic level.

Yuval: Do you get input from the pharmacist or the chemists, or is the group separate right now? How much integration is there, in terms of cross-functional integration?

Elica: Yeah, actually, we work a lot on being very integrated within the company and for the first several months we really went and spoke to more or less everyone from within BI that had anything to do with computational research in order to identify our use cases. And still for every use case that we work on, we do have an internal specialist, someone that is actually encountering this problem that we want to solve with quantum computers, in their work. And they work with us and they help us stay grounded. They help us define the use cases, the different, let's say, physical systems that we want to solve with the algorithms.

So, I would say that we are actually quite integrated within the company. We are the quantum computing lab and we are not part of, let's say, a bigger department or something like that. But we have put a lot of work into meeting the other stakeholders of the company. And I would say that we are quite well integrated already.

Yuval: From your experience, how much sharing is there between different pharma companies? I mean, sometimes when industries are young, people are saying, “oh, we are working together towards the same goal”. And may be, at some point later, they say, well, “now we're really competing”. Do you collaborate? Do you share information with other pharma companies in any significant way?

Elica: So, we have part of several pharmaceuticals, but not only also industry consortium that are looking at quantum computing. So, of course, there is sharing and we try to discuss what are the interesting use cases for us. But even though things are still pretty competitive, I think very soon they will start to be competitive. And, of course, all this work that is going behind the algorithms and the use cases, this is funded by our company or by somebody else's company. So, it doesn't make sense that you would share everything with everyone else, including with many that are interested, they might be curious, but they are actually not investing at all in quantum computing. So, it is kind of a funny situation where we do share, but overall this kind of, with the limited bandwidth, and of course, we also keep staff to ourselves.

Yuval: Other than more qubits with less noise. What else do you or the team needs to make progress to achieve even bigger business value?

Elica: And this is actually a very big goal, more qubits, also better connectivity, let's say. Because you can have many qubits, but also how they are connected. It also matters very much.

So, definitely we need the hardware improvement. Even if one day we have this, let's say, a few thousand error-corrected qubits quantum computers. Then we still don't know which algorithm to run on them because it would take such a long time to get the result, even on that computer. So, what we really need is still an improvement in the algorithms and kind of smarter and faster ways to manipulate the quantum information in order for that to become practical. And I think everyone that is serious about having a quantum computing value for their company and generating quantum computing value is really focusing also on the fault-tolerant algorithms development.

Yuval: When you design an algorithm for a particular computer and let's assume it was a gate-based computer. How worried are you about portability? Are you willing to make it hardware-specific or do you need to make it something that can be easily ported to something else.

Elica: Overall, I think it's safe to say that we are very much hardware agnostic, so we don't really care what is the particular hardware for implementing the qubits and the controls. So, from this point of view, we don't really care. And all the algorithms that we design can be implemented... For the moment, we are looking at the gate-based quantum computers. So, all of them can be translated to other gate-based architectures. In terms of having access to the hardware, this is not something that we think that it makes sense. Let's say, to buy a computer, a quantum computer, we always count on our partners to give us access to their hardware and to implement our algorithms and the use cases on their hardware, through the cloud.

Yuval: We are this time of the year that people talk about predictions. What are your predictions for quantum computing in 2022? What do you think will happen?

Elica: I am very much looking forward to seeing even bigger quantum platforms. There were quite promising expectations for this year for 2021, but I don't think we actually saw the promises already delivered. So, I suppose that this will happen in 2022. So, I'm looking forward to seeing some NISQ architectures in the qubit count above 100 Qubits. So, this will be really excellent. Probably also there will be some exciting news from the business world. We'll have other IPO's for sure. And also a lot more capital generated and funds raised, on the algorithmic side aspect also, because as I mentioned I know many industry players are already hiring and really building quantum computing teams. So, I also am looking forward to seeing some new algorithms for fault-tolerant quantum computers.

Yuval: As we get closer to the end of our discussion today. You mentioned that the team has five PhDs and when you grow it, I don't know if you're looking for additional PhDs. Do you think that's a problem? I mean, five PhDs maybe 5% of the eligible people in the world to do the kind of work that you're looking for.

Elica: Yeah. But, the truth is we are actually at the steady state and I don't think we will be growing much more in the next several years. Whether we will still continue to add the people with the same profile that team currently has, I think this is very debatable and it really also depends on what is the goal to add. What skill set do we need from the new more people. Maybe if we want to go more applied, even more applied, let's say, even more oriented towards the company. Then we can hire someone with more pharmaceutical experience and so on. But I think for building up a quantum computing lab, it really made sense to start by hiring quantum computing scientists.

Yuval: Elica, how can people get in touch with you to learn more about your work?

Elica: I'm very happy to connect with anyone. Mostly I use LinkedIn. So, they can find me on LinkedIn with my name Elica Kyoseva and they can send me a message or a connection request. And I usually accept those. So, this is, I think, the easiest way to find me. I also participate in some of the quantum computing for business conferences. I'll be happy if people come and share their views with me at the conferences.

Yuval: That's excellent. Thank you so much for joining me today.

Elica: Thank you Yuval.


My guest today is Elica Kyoseva, Quantum Computing Scientist at Boehringer Ingelheim. Elica and I talk about quantum computing in the pharma industry, the composition of their quantum team, the degree of competition and collaboration between companies, and much more

Listen to additional podcasts here

THE FULL TRANSCRIPT IS BELOW

Yuval Boger (CMO, Classiq): Hello, Elica. And thanks for joining me today.

Elica Kyoseva (Boehringer Ingelheim): Thanks, Yuval. Thank you very much for having me at your very cool podcast.

Yuval: Who are you and what do you do?

Elica: My name is Elica Kyoseva and I'm a quantum computing scientist at a German pharmaceutical company, Boehringer Ingelheim. previously, I was in academia and I have been working on quantum computing on several continents, including Asia, North America, and currently The Middle East, and based in Israel.

Yuval: Working in pharmaceuticals, I think there are a lot of use cases that people talk about why it could be fantastic for their future, for pharmaceuticals. But is there something that can be done today with quantum computing or do you think pharmaceutical companies are just preparing for distant future?

Elica: it's an excellent question. And the truth is that the biggest part of our effort is really to prepare for the distant future, which in our eyes is of course fault-tolerant error-corrected quantum computers. So, the majority of our use cases and really the problems that we are looking at are facing and answering the question of, will it be possible to implement these algorithms on the fault-tolerant quantum computer? However, even very recently, two weeks ago, we actually released a paper with one of our partners, QCware. And it is about NISQ calculations, where we show that using NISQ quantum computers can bring improvement to calculating binding affinity. So, this is really a step forward that is showing already tremendous progress for the NISQ era.

Yuval: How large is the team? 

Elica: It is not a secret. We are five people, which is actually pretty large. I know quite a few companies that are already ramping up and hiring as well. So soon, this will not be a big news, but we were probably the biggest team in the industries, especially in pharma. And all of us have PhDs and we all come from the traditional way of going through the universities, getting technical training, and then we kind of diverge. So, we have a person that was a postdoc at Harvard. I was personally even at the venture capital fund before I joined BI. We have IBM folks again in the university of Bristol. So, we are a diverse team, but I would say everyone is quite highly, technically skilled.

Yuval: When you talk about diversity, is it also diversity in education? I mean, putting aside where the education was. So, for instance, you have computer scientists and pharmaceuticals, or everyone has a computer science or physics major, what is the sort of structural composition?

Elica: So, from my perspective is it diverse, but it is diverse only from the quantum computing stack. So, we do not have people that are really, let's say, classically trained quantum chemists. The focus of our training was quantum computing. But basically it's zooming in on different parts of the quantum computing stack, all the way from algorithms and implementations and applications, but also all the way down to the hardware. So, personally, my background was in quantum control and I was looking at quantum optical problems of how can we control the qubits in a robust way. So, I was really looking at error mitigation of the quantum hardware.

Yuval: Do You think that full stack experience is going to be necessary in the future? I mean, I think you're quite unique that you work with the hardware and then the software and so on, but wouldn't people want to get more abstraction and say “ well, yeah, I understand what a qubit is, but I don't really care what the pulses underneath look like”.

Elica: Ultimately, I think it would be important to do both and software and hardware,  to meet in the middle. So, it is nice to abstract our algorithms and really to focus on applications. However, this can become quite detached from reality very quickly. Even if you think about many of the NISQ works and the NISQ algorithms, they are still noiseless. So, for many of these results, actually there are no noise models that are implemented on the system. So, ultimately it is very important for people to keep a connection to reality, and especially to the lifetime or the qubit and exactly of the different noise processes that are in the system and design the algorithms with that in mind. Of course, this is for NISQ algorithms for the ones that we want to see in the near term. For fault-tolerant, I think, there you can still be quite abstract and focus on the algorithmic level.

Yuval: Do you get input from the pharmacist or the chemists, or is the group separate right now? How much integration is there, in terms of cross-functional integration?

Elica: Yeah, actually, we work a lot on being very integrated within the company and for the first several months we really went and spoke to more or less everyone from within BI that had anything to do with computational research in order to identify our use cases. And still for every use case that we work on, we do have an internal specialist, someone that is actually encountering this problem that we want to solve with quantum computers, in their work. And they work with us and they help us stay grounded. They help us define the use cases, the different, let's say, physical systems that we want to solve with the algorithms.

So, I would say that we are actually quite integrated within the company. We are the quantum computing lab and we are not part of, let's say, a bigger department or something like that. But we have put a lot of work into meeting the other stakeholders of the company. And I would say that we are quite well integrated already.

Yuval: From your experience, how much sharing is there between different pharma companies? I mean, sometimes when industries are young, people are saying, “oh, we are working together towards the same goal”. And may be, at some point later, they say, well, “now we're really competing”. Do you collaborate? Do you share information with other pharma companies in any significant way?

Elica: So, we have part of several pharmaceuticals, but not only also industry consortium that are looking at quantum computing. So, of course, there is sharing and we try to discuss what are the interesting use cases for us. But even though things are still pretty competitive, I think very soon they will start to be competitive. And, of course, all this work that is going behind the algorithms and the use cases, this is funded by our company or by somebody else's company. So, it doesn't make sense that you would share everything with everyone else, including with many that are interested, they might be curious, but they are actually not investing at all in quantum computing. So, it is kind of a funny situation where we do share, but overall this kind of, with the limited bandwidth, and of course, we also keep staff to ourselves.

Yuval: Other than more qubits with less noise. What else do you or the team needs to make progress to achieve even bigger business value?

Elica: And this is actually a very big goal, more qubits, also better connectivity, let's say. Because you can have many qubits, but also how they are connected. It also matters very much.

So, definitely we need the hardware improvement. Even if one day we have this, let's say, a few thousand error-corrected qubits quantum computers. Then we still don't know which algorithm to run on them because it would take such a long time to get the result, even on that computer. So, what we really need is still an improvement in the algorithms and kind of smarter and faster ways to manipulate the quantum information in order for that to become practical. And I think everyone that is serious about having a quantum computing value for their company and generating quantum computing value is really focusing also on the fault-tolerant algorithms development.

Yuval: When you design an algorithm for a particular computer and let's assume it was a gate-based computer. How worried are you about portability? Are you willing to make it hardware-specific or do you need to make it something that can be easily ported to something else.

Elica: Overall, I think it's safe to say that we are very much hardware agnostic, so we don't really care what is the particular hardware for implementing the qubits and the controls. So, from this point of view, we don't really care. And all the algorithms that we design can be implemented... For the moment, we are looking at the gate-based quantum computers. So, all of them can be translated to other gate-based architectures. In terms of having access to the hardware, this is not something that we think that it makes sense. Let's say, to buy a computer, a quantum computer, we always count on our partners to give us access to their hardware and to implement our algorithms and the use cases on their hardware, through the cloud.

Yuval: We are this time of the year that people talk about predictions. What are your predictions for quantum computing in 2022? What do you think will happen?

Elica: I am very much looking forward to seeing even bigger quantum platforms. There were quite promising expectations for this year for 2021, but I don't think we actually saw the promises already delivered. So, I suppose that this will happen in 2022. So, I'm looking forward to seeing some NISQ architectures in the qubit count above 100 Qubits. So, this will be really excellent. Probably also there will be some exciting news from the business world. We'll have other IPO's for sure. And also a lot more capital generated and funds raised, on the algorithmic side aspect also, because as I mentioned I know many industry players are already hiring and really building quantum computing teams. So, I also am looking forward to seeing some new algorithms for fault-tolerant quantum computers.

Yuval: As we get closer to the end of our discussion today. You mentioned that the team has five PhDs and when you grow it, I don't know if you're looking for additional PhDs. Do you think that's a problem? I mean, five PhDs maybe 5% of the eligible people in the world to do the kind of work that you're looking for.

Elica: Yeah. But, the truth is we are actually at the steady state and I don't think we will be growing much more in the next several years. Whether we will still continue to add the people with the same profile that team currently has, I think this is very debatable and it really also depends on what is the goal to add. What skill set do we need from the new more people. Maybe if we want to go more applied, even more applied, let's say, even more oriented towards the company. Then we can hire someone with more pharmaceutical experience and so on. But I think for building up a quantum computing lab, it really made sense to start by hiring quantum computing scientists.

Yuval: Elica, how can people get in touch with you to learn more about your work?

Elica: I'm very happy to connect with anyone. Mostly I use LinkedIn. So, they can find me on LinkedIn with my name Elica Kyoseva and they can send me a message or a connection request. And I usually accept those. So, this is, I think, the easiest way to find me. I also participate in some of the quantum computing for business conferences. I'll be happy if people come and share their views with me at the conferences.

Yuval: That's excellent. Thank you so much for joining me today.

Elica: Thank you Yuval.


See Also

No items found.

Start Creating Quantum Software Without Limits

contact us