Tomorrow's Medicine, Ep. 15: Wearable Ultrasound & Career in MedTech - Dr. Mai Elfarnawany

With Dr. Mai Elfarnawany, VP of Technology, Flosonics Medical. Hosted by Arseniy Arsentyev. Published June 9, 2026.

https://aarsentyev.com/podcast/ep15-mai-elfarnawany

Machine-transcribed and lightly edited: filler words and false starts are removed, and the wording is otherwise as spoken. Where the two differ, the recording is the record. Times are minutes into the recording. Not medical advice.

[0:00] Arseniy Arsentyev: Hey everyone, it's Arseniy and it is the 15th episode of Tomorrow's Medicine podcast. Today I'm talking to Dr. Mai Elfarnawany, who is the VP of Technology at Flosonics Medical. Dr. Mai, could you please introduce yourself and maybe tell us more about what you do?

[0:18] Dr. Mai Elfarnawany: Yes, absolutely. I'm delighted to be here and to be able to speak to your audience. I'm Mai Elfarnawany, I'm a biomedical engineer by background. I have a PhD in biomedical engineering with expertise in ultrasound. And throughout my career, I initially started as an engineer and then grew into more leadership and management roles. And now I'm serving as the VP of Technology, where I oversee the engineering, software, hardware, firmware, data engineering, as well as the IT departments. And we are building FloPatch, which is the first wireless wearable Doppler ultrasound.

[1:08] I'm sure we're going to talk about it a little bit more. So I lead a team of 20 engineers today and it started off with being me first as the first engineer on the team, the first software engineer on the team.

[1:25] Arseniy Arsentyev: Before diving into the story of FloPatch, could you tell us how you first joined the Flosonics Medical after the university? Was it your first job?

[1:37] Dr. Mai Elfarnawany: No, I did. So as I mentioned, I did a PhD. So first I was completely in academia. I started off as a research scientist. That's what I had a vision for since I was younger. So I did a PhD followed by two years of postdoc studies. And then there was a point where I realized that academia is a bit slow paced for what I'm hoping to achieve. And so I started looking into the transition into industry. With a PhD and a postdoc, with no prior experience in industry, it is a rougher transition. And so I had to, I call them bridge jobs, where I was joining more of a, like a technical expert software developer for a couple of years before I joined Flosonics.

[2:32] And those bridge jobs were really, really helpful to up my technical skills and make me up to the level that the market is needing. And then I joined Flosonics seven years ago now, when the team was just a team of 10. And then I grew with the company as it grew.

[2:49] Arseniy Arsentyev: So we can say that you're with Flosonics from its beginning, and you actually grew the whole technology department on your own.

[2:58] Dr. Mai Elfarnawany: With the help of a great team. But yes, I was a key player in the growth of the tech team.

[3:06] Arseniy Arsentyev: All right. And everyone now wonders, what is this Flosonics company? What is FloPatch? And for me, even given that I do some research and reading much in medicine, it was new that there are some types of devices like a FloPatch. And could you briefly describe the key idea of what it is, why some patients need it, and why it is the future? Because I can assume that there's a future due to the fact that you are working on it.

[3:40] Dr. Mai Elfarnawany: Absolutely. So Flosonics Medical is a startup, and it is focused on the one product currently, which is the FloPatch. FloPatch is a wireless wearable ultrasound device. So unlike the ultrasound devices that you drag them in a cart, and then you use them for multiple purposes, it is an ultrasound device that is basically worn on the neck with an adhesive. It is wireless and that is also a difference in comparison to typical ultrasound devices and its placement is very fixed so that it can continue to collect data in a very reliable way. Whereas typical ultrasound would require the user to actually place the probe and capture the signal.

[4:28] So as you can think, the ultrasound devices have so many applications. They can be multipurpose, they can be used in so many use cases. And similarly, FloPatch, because it is an ultrasound technology, it can be used in many, many applications. However, for a startup, for a commercialization of a product, we had to focus on a specific use case, which is the sepsis use case. If you're familiar, sepsis patients would come into the hospital and decisions need to be made very quickly as to whether to provide them, give them fluids or not give them IV fluids.

[5:10] And one of the key things that FloPatch helps with is determine whether this patient can tolerate, this patient is going to be responding positively to those fluids or not. And it is really critical because if a patient's heart cannot tolerate the fluids that they get, then their system and their body starts to crash. It's called a fluid overload. And I'm not sure if you're familiar with people going into the hospital and then everything getting swollen, like their limbs would get swollen, their face would get swollen. And that's basically a fluid can be caused by fluid overload where the heart couldn't handle the fluids. So FloPatch's main target market today is to be used in the sepsis use case.

[5:56] The clinician or the nurse would basically stick the FloPatch on the patient and then use the app that we provide and walk through specific steps to assess how the patient's body can respond to fluids. And we basically simplified the workflow very, very, very, very much to tell yes, it is likely to respond or no with specific clinical metrics that we have developed. And so it would make it much easier and more reliable for the clinician to make that decision. Yes, you can give this patient more fluids or let's avoid giving them more fluids because that can risk a fluid overload.

[6:35] Arseniy Arsentyev: General patient imagines ultrasonic device as this huge box of advanced equipment that is supposed to provide information and verbal ultrasound sounds almost sci-fi. So why is it possible to do this small adhesive devices now and not turning? I think you started actually 10 years ago, not like 30 years ago. What changed now and what perspectives on the market can you see?

[7:09] Dr. Mai Elfarnawany: Of course, there is a lot of a lot of advancement that allowed us to get to this point. The first one is the ability to actually have a very reliable wireless transfer channel that can pass a huge amount of data that then gets processed and then presents the metrics and the clinical metrics that we would be needing to make decisions. The miniaturization of the boards and the hardware is also a very important component because before early, early days, what we are able to do with FloPatch today needed a much bigger footprint and electronics design to achieve that.

[7:53] For us in particular, FloPatch is not like I have to say, it's not a general ultrasound device. It doesn't do everything a focus device would do. It is mainly a Doppler ultrasound device. So it has a very targeted mode of ultrasound that is achievable with this miniaturized and wearable technology.

[8:15] Arseniy Arsentyev: How deep can a scan?

[8:18] Dr. Mai Elfarnawany: It goes up to four centimeters under.

[8:22] Arseniy Arsentyev: Four centimeters. That's significant, honestly. And the next question for you is, given, let me formulate it, given this ultrasound device's portability and that is not, as I understand, it's not constantly connected to some source of energy or something. What are the troubles of manufacturing it and how your Department of Technology, Dr. Mai, is related to the manufacturing process? Because I've seen many podcasts where people speak about their device companies and the huge part that is often overlooked is manufacturing process. How do you deal with it in Flosonics Medical?

[9:17] Dr. Mai Elfarnawany: We actually have our full manufacturing in-house. We have developed every step of manufacturing from the sensor to the port.

[9:26] Arseniy Arsentyev: Sorry to interrupt you, Dr. Mai. Are you located in Canada?

[9:29] Dr. Mai Elfarnawany: We are located in Canada, yes. Our office is in Toronto and we built the whole device in-house. Of course, we get some components, sub-components of the device built in third party and then we get them shipped and then we do the assembly. But the key component in our device, which is the sensor, the ultrasound transducer is built fully in-house. And I think that has given us a lot of strength and agility in how we have developed the manufacturing process over time.

[10:04] We are the experts in FloPatch and so if there's any refinement or improvement in the manufacturing process, we were very adaptable and capable to adjust the production process to produce high-yield devices.

[10:21] Arseniy Arsentyev: And speaking about growing your team, could you kind of reiterate on those moments when you started alone, as you said, in the engineering department and started to assemble the first team? What challenges have you faced? What have you understood? And is this axiom correct that working in a team, working in a huge team and working alone are three completely different mindsets and completely different skill sets was the most important?

[10:56] Dr. Mai Elfarnawany: Absolutely. So I mentioned I joined as the first software engineer. There were engineers for the hardware parts of the device as well as the transducer. We have an expert transducer scientist in our team who has been with Flosonics since it started, which is a great asset. I was the first software engineer and I built a software team. The growth, of course, between what I was doing and how my day was when I first started is completely different than what I'm doing today.

[11:30] I was writing the first algorithm that went into the device. I spent days and days debugging the code and documenting the code. So it was a complete change to what I was doing. And so from an experience perspective, being an individual contributor is drastically different than being a team leader. But I've enjoyed both and I still enjoy both. And building the team, I think you were asking about the challenges or some of the moments that were transformational in our growth. There's some interesting stories from the time. Yeah, so I have to say it was really my first experience hiring when I was promoted to be manager.

[12:19] And I would say that initially with a small, small team trying to build the first functional prototype of the device, really what we were looking for are people who can do a lot of things. Like in a startup environment, you want people with multiple hats. They can do software. They can do coding. They can do testing. They can do it all. So that was a very different profile when hiring than what I'm doing today. Today, my team is 20 people. They are all experts in what they do. And so as we grew, the needs for the team and for the technology department had to become more focused and more expert required. And so we were really hiring an expert in cloud infrastructure, an expert in ML, an expert in AI.

[13:10] And that's really, and having that flexibility and understanding that the different stages of the business required different profiles of people to join the team was really, really crucial to help us have the best team at all time points in our journey.

[13:27] Arseniy Arsentyev: Dr. Mai, a geeky question, since I'm a bachelor's in computer science. And what were your job like? Were you writing an assembly code for this device? I'm not sure if you're allowed to share this technical details, but what type of programming are we talking about? Are we talking about high level engineering when you connect Python cloud services or was it really in detail in a nutshell?

[13:55] Dr. Mai Elfarnawany: No, no, it was more embedded level. To achieve the performance required for processing, we really needed to write our algorithms in a low level language so that it can keep up with the processing requirements and the performance that is required on a small chip on a device and then to pass all of that information to the iPad which is basically where our software lives. It required a low level of language so when I first joined I was more developing the algorithms and really with Doppler signal or with ultrasound signal there's a lot of signal processing that needs to go in to allow for the...

[14:37] It's a very rich signal like I call the Doppler signal a three-dimensional signal there's everything tells you about the velocity of the blood it tells you about the intensity of the blood which is how the density of the red blood cells. It also changes with time, it can capture pulsatility and heart pumping so it's very very rich and so my first job was to try and extract the useful information from that rich data and that was basically writing different algorithms, different signal processing algorithms.

[15:11] So yeah and then I had to write them in a low level language, not to the level of assembly, one step up.

[15:18] Arseniy Arsentyev: All right and now you have this platform at FloPatch, is it only... So you have a device for a patient that is adhesively just attached to the person's body. And you have this iPad, as you said, iPad, mobile app. Do you have something similar for the doctor side of the equation? Is FloPatch a whole IT platform or is it for now only the mobile apps?

[15:55] Dr. Mai Elfarnawany: It is a bedside device. So the monitor is there. The clinician uses the app to see the data as it's coming in, to see the metrics and the analytics that we provide in the application. So it is a bedside device that gives real time results. We do have the capability and we have built the infrastructure for a whole cloud backend that can, in the future, provide more aggregated data and reporting for the different hospitals and customers. But that is not our current focus. We're still nailing down the real time bedside workflow so that we can, again, as I mentioned, our target is sepsis. And with sepsis, you really need to have an immediate answer. Like our competitors, the devices that can give a similar answer would require a much longer assessment duration.

[16:54] FloPatch can give an answer within three minutes maximum, which is necessary for dealing with sepsis cases that are critical.

[17:03] Arseniy Arsentyev: Can we say that your company's biggest innovation is speeding up of the process of ultrasound analysis or the sepsis patients?

[17:16] Dr. Mai Elfarnawany: I would say speed it up and simplify it.

[17:20] Arseniy Arsentyev: Simplify it?

[17:21] Dr. Mai Elfarnawany: There's not a lot of components that need to be connected to the patient. There's no manual placement requiring an expert person to know exactly where to place it and how to gather the information. It is for the nurse user. And it is simple, put it, attach it, then you can really easily follow the workflow that we have provided and get an answer within three minutes.

[18:32] Arseniy Arsentyev: So I was talking about, do you envision the world where the majority of people would use ultrasonic sensors and devices, like right now is happening with constant glucose monitors, people who are trying to defy the aging and to, we can say max their health, are using glucose monitors for different reasons and for different workouts. Do you see something similar happening with ultrasound devices?

[19:04] Dr. Mai Elfarnawany: I think from my perspective, I think the ultrasound is still, it is too rich of an information pipeline to be general facing the public in terms of benefiting directly from it. I think with FloPatch and similar technologies, when ultrasound is very focused and is providing a very specific output, I believe that, yes, it can scale to become something that can be broader, used by the public to assess health indicators. In cases like this, I can envision a future where FloPatch, people would put it on to, for example, assess how their heart is functioning after a surgery or around a specific time before a sports event and they want to assess how the blood flow is stronger or affected before a big sports event.

[20:15] So I can envision applications, but I think because of the richness of the data, these applications have to be very specific and the user-facing components of the tools would have to also be specific to that use case. But imagining that we just give FloPatch to people and tell them, take it and track your heart functionality without more focused output will be harder for the general public, I believe.

[20:47] Arseniy Arsentyev: And what's the battery life of the device currently?

[20:50] Dr. Mai Elfarnawany: Currently, we have it can have streaming up to two days. And typically the use case with sepsis is that it is an assessment. So they put it on and they measure and then it basically powers down. If they want to do another assessment, they turn it back on. So it can allow for multiple assessments per patient. Basically every patient gets a device and they can use it for multiple assessments over time at entry and then two hours after. And then they give them a treatment and they want to assess their fluid responsiveness. Again, they can basically use it multiple times.

[21:28] But for the general monitoring, like if you're asking about using FloPatch for the monitoring application, we have the capability, we've used it in some surgeries like CABG and it was monitoring throughout the whole surgery. So it has the capability for it. But as I mentioned, it's so much data that in order for it to be very useful, we have to focus it and give some metrics or clinical metrics to the user that is using it for monitoring.

[21:59] Arseniy Arsentyev: So the actual device is just a tunnel for the data. It transmits data. That's why it's so energy effective. And where does the computing take place?

[22:12] Dr. Mai Elfarnawany: All the computing happens on the iPad. We've leveraged all of the iOS operating systems and chips for our high intensity, high complexity processing.

[22:26] Arseniy Arsentyev: As you said, the complexity of the signal from the chip is enormous. It's a 3D landscape with velocities with different parameters. Is this iPad's performance enough or do you feel that maybe it's a great time to transition to more powerful computing techniques like the cloud or having your own servers in Canada? You did the manufacturing, so servers will be definitely a step easier.

[22:54] Dr. Mai Elfarnawany: We have, so we are continuously monitoring our performance on the iPad. And what I can say is like as new versions are coming out, it's actually providing stronger processing ability. Now the chips can handle ML and AI processing seamlessly. We also have a very, very strong software development team that writes very optimized code and optimizes it consistently. So our current system and operating systems in the iPads are sufficient for what we need. We haven't needed to shift yet.

[23:36] Arseniy Arsentyev: You're the first VP of Technology on the show, and it's my privilege to ask these questions that all young people are curious about. Do you use AI in your workloads inside the company? How much do you use it? How is it integrated? Because many young people who do startups right now just vibe-code like 90% of the stuff and somehow it works for now. What's the AI policy in your company?

[24:06] Dr. Mai Elfarnawany: So we have integrated machine learning as needed in our pipeline. So we have, today we have hit using FloPatch on 10,000 patients. So we have an enormous data set that we have leveraged machine learning to allow us to more accurately measure some clinical metrics using the data that we have. And so I wouldn't say we have an AI integration as in AI taking over our pipeline from start to end. We still have a very engineered pipeline for signal processing. And we have only integrated machine learning in specific places where we found that logic-based algorithms or analytical algorithms are not performing as well as ML-based algorithms. So it's never like an end-to-end ML replacement. It is just a particular component in our pipeline. I think that has given us strength for best use of ML.

[25:20] Because ML and AI are very popular now. And the underlying determining factor for a successful ML and AI or not would really be the data. If the data is not good, or the data pipeline or the infrastructure to clean up the data and preprocess it and then integrate it, it's never going to produce a successful algorithm. And so with this focused approach of injecting models in specific areas, we are really confident that the data going in is up to the level that it requires. And so it does that functionality at a very high performance level.

[26:10] This is for our product, like AI and ML in our product. I have other points about where else we have used AI in the company for the technology department. Of course, we have used it in other places. But for the product, we have been careful to only integrate AI and ML in the right spots with a very focused approach.

[26:33] Arseniy Arsentyev: Do you see yourself integrating into your agile cycle, the agents and hiring them, hiring the agents?

[26:45] Dr. Mai Elfarnawany: Sorry, in our what?

[26:48] Arseniy Arsentyev: In your development agile cycle. Do you see yourself, as I said, not working with people, but working with the vendors of AI services and completing certain programming, we can say, tasks with the use of AI engines?

[27:09] Dr. Mai Elfarnawany: So I can't predict the future. I think there will be a time in the future that that will happen, where some roles would be efficiently substituted by an AI agent or automated by an AI agent. But with the medical device field, I think that will be a bit farther down in comparison to other industries because of the criticality of the medical device algorithms and processing. Like we, in order to release a small change in our software, it takes us some good time to do all the validation, verification, provide all the reporting. And so I would say it's not going to be a very simple plug and play, take that role out, put that AI agent in because of the criticality of the output of these algorithms.

[28:07] So I still see, what we have done is basically we have provided all our developers the accessibility to using AI agents to support their work, to give them some heads up in developing a component in our stack. But then there's always that expert developer that is reviewing, ensuring it is doing what it's doing, confirming that it is performing up to the spec and then typically documenting that in a report. So, yeah, I think one day in the future, maybe less or slower on the med tech side because of the criticality and the impact. Like one error in code is not just a website being down, it's the health of the patient. So it is more critical.

[29:01] And the FDA and all the regulatory bodies have provided, and continue to provide guidelines on how to integrate AI so that the patient safety is at the top priority. So, yeah, I think one day, but today we have leveraged it to the best of what we need.

[29:23] Arseniy Arsentyev: That's a great point to transition into a huge topic I wanted to talk to you about, which is the policy, the compliancy of engineering departments in our healthcare space. I know that you're governed and you're sometimes restricted, sometimes safeguarded by many rules in healthcare policy because you're working with real patients data. Do you have any insight how to straighten this process for future med tech founders, for future med tech engineering teams? How to make this process as simple as possible and always stay compliant and ethically compliant, as we can say?

[30:12] Dr. Mai Elfarnawany: This is a loaded question. It is basically multiple years of experience to get to a point where you feel comfortable and yet you're not too comfortable. Because, again, the purpose of these rules and guidelines is to keep the patient safety first. So we always respect it. In terms of learnings along the way that made things easier for us to maintain compliance, I would say documentation is the number one component. Engineers don't like documentation as much. Like they like solving problems and not writing about solving them. So I would say the biggest thing then, and it's something that whenever I was hiring someone to join the team, I would tell them, we have drastic amounts of documentation required with every change you make. And so are you okay with that? So I would say, yes.

[31:10] The first thing is to document along the way and not right before release or right before submitting an update. That is really crucial. Also automation. For example, we have built in-house testing platforms: even if it is not necessarily the QMS-level testing or V&V requirements, we have built additional, we call them internal informal, testing. And that is really important because it helps us continuously test all the aspects of our product as we are developing. So we have these automated robot frameworks that does testing end-to-end, like the full stack from hardware all the way to uploading stuff to the cloud. And that is running all the time. It's like on a weekly basis, we have different triggers.

[32:04] So we have invested a lot of effort in creating these automated testing suites that give us the confidence that if there's something that is not following the required specifications, it will be picked up early, early, early time, not after we release. I would say I'm very proud of our team that whenever we release a new version of the software, we have really rarely needed to push out a patch release where you basically do bug fixes, you realize that there is a bug and then you have to push out. We have to maintain a very steady schedule of releases because of this ability to do a lot of testing ahead and a lot of dry runs ahead and catch everything before it goes out.

[32:53] So I think these are key components to maintain compliance: document, because it makes the submission of the new update much, much easier and simpler, and making sure that the product is functioning as it is required all the time. And not just right before release or right before submission. The third component is we have people who are dedicated to always review all the regulatory requirements. If there's anything that gets pushed, we don't wait until we are pinged about it or there's an audit. We always try to implement that in our workflow as soon as it comes out.

[33:31] And so that also saves us a lot of headache when we are about to do, when we are about to get an audit or when we are about to do a release that we need to do something quickly or test something. But we have already integrated it in our development cycle earlier.

[33:48] Arseniy Arsentyev: And how often do you have this audits, the code base audits? Do people from FDA or other regulatory bodies actually review the code?

[34:03] Dr. Mai Elfarnawany: I think from my experience, they have access to the code, of course. We give them access to the repositories. But as I mentioned, we have documented our design details really, really well. We have even provided them with exports from our code base that they can see all the classes and everything is laid out in a document. So I don't think we have had an ask to go and review our code basis. Rather, they just review those documents that we provide the output and the design details. And it provides them with sufficient information to know all the ins and outs of our design.

[34:43] For audits, it depends on the organization. So FDA, they have specific timelines for audits. We are also EU certified, so we get audits. All the dates are not top of mind, because we have an amazing regulatory team that takes care of all that. But I think the EU gives us audits maybe annually, once a year or once every two years. I'm not really sure. But yeah, there's definitely regular reviews that happen to our QMS portfolio as a whole.

[35:22] Arseniy Arsentyev: And can you take me back to FloPatch version 1.0 or I don't know how you named it and what changed? How can you describe the enormous progress that happened since that time?

[35:38] Dr. Mai Elfarnawany: Very, very interesting. I joined the company after they had a good functional hardware prototype. So I know version one was literally a huge box with a wire. It wasn't wireless to start. That was version 1.0. And I think the first FDA version that was cleared was wireless, but it was maybe double the size that we have today. It was just a big box. So now FloPatch is literally handheld.

[36:12] Arseniy Arsentyev: Everyone, look at the video feed if you're listening. That's FloPatch.

[36:17] Dr. Mai Elfarnawany: So it is very miniaturized. It's completely wireless. There has been a lot on the hardware in terms of optimization, a lot in the firmware improvements to make it very user-friendly, like all the power downs and the battery and the power management has improved drastically from version one to today, to allow it to be used that extended, as I mentioned to you. Of course, for the software, I was the first one to write the first algo of FloPatch and it was working, but not really the way it is today. It's much, much better in terms of the signal, extracting the signal characteristics in terms of making all the calculations. So I would say over the last seven years, the software stack in itself maybe has transformed three times from start to end.

[37:13] Every single component in the software stack has changed to the better. We have integrated, as I mentioned, ML models at some point that made our performance, basically brought it from a 70% performance and accuracy to an 85%. We're at now 95% accuracy, which is really, really high. And it is only attainable using these advanced and complex models. And then the wireless channel, like when I first joined, we were, when you get the data, it's like if you have a 16-bit data, we were only able to pass 8-bit in the Bluetooth channel. And we were able to develop algorithms that allowed us to pass the full 16-bit. So we got the complete data through and that also leveraged our ability to extract information much more reliably and fully.

[38:11] Arseniy Arsentyev: And the people who are motivated by your speech today and maybe want to join the Flosonics Medical, what do you look for whenever you're hiring engineers? Is this a math-heavy job that you're working with the signal processing, as everyone understands, I hope I know? And does it require advanced math, special skills that are not usually taught in the university? And what is the average programming stack of your development team?

[38:44] Dr. Mai Elfarnawany: Very, very loaded question, too. So for my engineering team, it's really wide net now because it's not only the algo team. We have an algo team, we have a cloud team, we have an iOS team who is basically developing the application. I can say that as we shift more into ML, I think the future is in integrating more machine learning models into the pipeline to make it more robust and to advance its use cases. As we move from sepsis to other use cases, I'm sure we will need the people who are able to bridge both worlds, the scientist or the research mind and very good development skills. That's really the core, and we call it actually the processing core.

[39:38] Our core team needs to have that. They are all PhDs and they are all people who can write code to an industry level. So that would be for the core team. The cloud team, I think, requires people who are front end, full stack typically, and have experience in machine learning and Kubernetes and basically maintaining an infrastructure for the cloud. So it is a wide scope of skills and it depends on what role people would be interested in joining. But if it is the core team, I think it is not math heavy. I would say research, research mindset, plus exceptional coding skills. Our stack has everything from Python to Angular to C to C++, Swift, and our developers write all of these languages, basically.

[40:40] Arseniy Arsentyev: Let's proceed to the final part of this interview, which is, we can say, the second life of Dr. Mai. You're speaking about women in med tech, in technical jobs. You're a career tutor and you're talking about your career. The first question on the subject is, if I'm counting correctly, four or five position upgrades, four or five forks during the latest years. How is it possible? And what's your secret?

[41:19] Dr. Mai Elfarnawany: Thank you. This is a great question. What's my secret? I would say that growth is really internal before it is external. For my career journey, I would say it requires a very good environment. So the actual company that you're in will really bring you higher or keep you stagnant. So first of all, know where you're joining, who are you working with? And I think that is a key component. But after you have found a very good environment with people who are supportive and encouraging.

[41:58] I have basically every step in my profile was backed with something that I worked on myself. So there was a point I was just a technical individual contributor. And then in order to become the manager, I basically invested in understanding all of the regulatory requirements, all of the documentation needs. And then I was able to kind of guide and support the other team members to produce code in a way that is very aligned with the regulatory requirements. For me, luckily, every step up in my career, I had done that job before I got the title update. So I was managing the team and supporting them to document their work and write the code in a level that is FDA compliant before I became their manager.

[42:59] And then when I moved from manager to director, I basically had done a PMP on my own. I found that in order to efficiently manage the team and manage multiple projects and being able to supervise a lot of moving parts at once, I needed some tools, I needed some techniques. And so I went to look around and found that having a PMP will help me with that. I was able to bring in project management tools in the team. I was able to integrate it and to allow people to work efficiently, more productively, being able to deliver on their goals.

[43:40] And then before I got into the VP level, I really had to, I was very curious. I was purely an engineer, but I was so curious and asking so many questions about strategy and what are we doing with business and what's the fundraising? Tell me all about it. I would like to know about the commercial plans. And so I think in order to grow, you need to be growing internally first, and then you'll be able to find a place where this internal growth is actually leveraged externally too.

[44:15] And similarly with my recent promotion from VP of Engineering, which I've had for two years to VP of Technology, the main transition was realizing that we are a huge team now. Like we're 80 people and we don't have an IT department and IT is becoming a bit of a headache for everybody. It's not really helping everybody to not have an IT function. And so I took the initiative and said, I'm going to do hiring for an IT person. We're going to build a whole IT stack, a whole IT system. I'm willing to supervise it. I'm willing to oversee it. And so then I basically expanded my scope from only engineering to being able to see all of the technology, all technology functions.

[45:04] So yeah, I think that's my secret is grow internally, do what you want to be doing after first, and then it will be natural to grow into that function.

[45:20] Arseniy Arsentyev: And how not to stop? How not to stop progressing? And at one point, not to say to yourself that maybe it's enough, maybe I need to stick where I am for like five years. We have this hard coded into our brains from our ancestors, this time scapes that it used to take 30 years to get a promotion. Like I'm exaggerating, but it's close to the truth. And how to switch your brain to constantly growing. That's what I'm asking you.

[45:58] Dr. Mai Elfarnawany: So are you asking if I'm understanding your question correctly, you're saying we are naturally wired to prefer stability?

[46:06] Arseniy Arsentyev: Yes, absolutely.

[46:08] Dr. Mai Elfarnawany: And how to switch from that mindset of wanting stability to wanting growth?

[46:12] Arseniy Arsentyev: And we have this notion of our capabilities. For example, let's say I'm working as an intern in finance. And I'm psychologically based on this level of an intern. And even if my skill set, knowledge set, experience set is higher, since I'm holding this position, and I'm just starting my first job, I'm on this level. And how to break the ceilings and actually tell yourself: no, look, I can do more, I can lead this, I can manage that, and I can put it all together?

[46:56] Dr. Mai Elfarnawany: It's a good question. I really never thought of projecting it on myself and asking. I think the way to break that ceiling: for me, I never use the title to define what I do. My measure of what I can do is my skills. So even if my title is here, as I mentioned to you, I was a software developer, but I have a skill of managing projects. So I manage projects. I don't make the title hold me back. And I don't necessarily put them side by side. Here is my job description. Here are the things I'm doing. I'm not going to go above any of that. Because the truth is, to grow, you need to go, you need to hit your boundaries, you need to go out of your comfort zone.

[47:50] And that whatever is listed in the job description is your comfort zone is what you signed up for. But in order to grow, you need to do more. You have to push against that. You have to move one step outside the comfort zone. You have to do one thing that you didn't do before. You have to try one thing that you're not sure you can do so that it results in your growth. If people remain in their circles, they remain in their comfort zones, they will function within that. But if they're interested in growing, and it has to be a personal interest too. Some people prefer just to remain where they are, and that's absolutely fine.

[48:24] But if someone wants to grow, and they don't know how to do that, this is the key to break the wall, go out of your comfort zone, try something they haven't tried before. And I have to say, I forgot to mention that, but one key component to my growth was not internally driven only, but also to have mentors. I had mentors for every single stage of my career. So initially, I had an ultrasound scientist as a mentor, where I would bounce ideas about the algorithms that I was developing with. And he would guide me. He's like, no, this is not attainable. This is not achievable on a small chip. Maybe this is too complex. I had a mentor at the technical level. I had a mentor for project management.

[49:14] I had a mentor who was an amazing woman engineering leader in a bigger company. And then I had a mentor for how to transform a company from working in a small scale to being a large team and align all the goals and project manager at a higher level. So internal drive, plus skills, hit the boundaries and have a mentor. If we don't have a mentor, we will still grow, but it will be more bumpy, because you're not benefiting from the experience that someone who has done it before has provided you.

[49:55] So they would tell me avoid this or do this. And that would definitely make the growth much more linear than a very bumpy, bumpy step up.

[50:05] Arseniy Arsentyev: And how do you find mentors? Was it your was it in your case that you were surrounded by them naturally, since you're working for this company and this community and this team? And how people who do not have obvious mentors on their job can find one?

[50:23] Dr. Mai Elfarnawany: Very great question. I think it was a combination of reaching out. So I would reach out to my teams. Does anybody know someone who is an expert in Doppler ultrasound that I can talk to? And one person will lead to another to another to another. So in the cases of the very, very expert mentors, it came through the network. And if, again, you just need someone who has done it before, and it doesn't have to be the top notch scientist in that field to be the mentor, like someone who has that will definitely be one step from where you are.

[50:59] The other approach that I have leveraged for mentors is really joining mentorship programs. There's a lot of virtual mentorship programs out there. And I would typically I remember like one month, I would just go and sign up to all mentorship sessions as a mentee. And then you meet mentors there. You meet people who are passionate to guide you for your career for a med tech, for example. I would apply to mentorship programs. And there I would meet someone who is, again, a woman leader or someone who has transformed their business using an amazing system. And then after that session, I would get their LinkedIn, connect with them, schedule a one on one, talk to them about it and ask if they are willing to do like a monthly voluntary mentorship session.

[51:50] I have never met anybody who said no.

[51:55] Arseniy Arsentyev: And a tough question for you. A very tough one. And from your experience, do these mentors participate in mentoring other people purely from their desire to help people, to bring something to the table, to help the community, to help to see you grow? Or have you seen that some mentors, because I did, want this mentoring to be a resume checkpoint: yeah, I'm doing mentoring, I have people, I'm helping them, and not fully being into the process of mentoring?

[52:38] Dr. Mai Elfarnawany: I've seen both. I've seen both. But even in the case of a mentor who is really using me as a point in their resume, if you know what you want to get out of them, if you go into that relationship with some questions, and they have answers for it, you will still get the answers. Whether that is something that will continue. Sometimes I had mentorship relationships that lasted two months, and then it died down because of something similar to what you're describing. Other times I had a mentor for two years. And that was someone genuinely wanting to help. So I think whether it's this or that should not hold back from, trying to leverage that relationship, because you will benefit either way.

[53:21] Even if it is just a point on their resume, you'll still get some answers. And that will be beneficial.

[53:31] Arseniy Arsentyev: Finally, could you tell me about the work you do to support women in tech? And what are the current problems on the landscape?

[53:41] Dr. Mai Elfarnawany: So for me, I've always participated in women in something. So initially, when I first graduated, I was in university and on campus, joining all of the women in engineering programs. I find that women in engineering and women in tech in general, there's the challenges are really being different than the majority of people in the space. And so with difference, some challenges come, such as needing to prove oneself, needing to learn how to speak about yourself and about your skills, how to voice your opinion in a room full of men. So I've had the passion since I graduated, basically, to find in every stage that I'm in, a target group that I would be able to benefit. So it started with women in engineering. I helped out there. And then I started joining.

[54:40] There's an amazing organization in Ontario. It's called OBIO, where they are really focusing on women in tech and med tech in particular. And so I joined their networking events. I joined their breakfast series. I joined. And really, I tried to connect and be part of that network, which then leads me to have, for example, there's a number of mentorship activities that I do today for women CEOs of startups. And I really met those people in those events. So being there, being out there, being present and available allowed me to have more meaningful mentorship relationships with women in tech.

[55:25] I've even gone to California with another organization that I do speed mentorship with. And again, it was really, really helpful as I met some other students there and was able to talk to them about their career and career growth. So I think my strategy is to look around, look at women in tech platforms, join them in whatever capacity as a speaker, as an attendee. And from there, I will be able to connect with people who need my support and I can help with whatever I've learned along the journey. So I think it's really helpful.

[56:07] Arseniy Arsentyev: I know so many brilliant women who are experts in their fields and I personally do not, maybe it's due to the community, I do not notice some biases or some things that are currently on the air. What can you say, how severe is the problem? Is it getting better in Canada because you're living in Canada and in the US? Is it getting better and is it now a better place, IT is a better place for women to work in?

[56:43] Dr. Mai Elfarnawany: I think, of course, early days it was much harder because of the, at least because of the numbers, like the early days, the number of women engineers would be much less than men. Today, I think there has been a huge growth in the numbers and interests and skills and expertise. I'm seeing a huge booming in entrepreneurs, women entrepreneurs, knowing the problems that are like the femtech, for example, is an amazing field that I think just emerged a few years ago. All the technology that is focused for women needs is an amazing switch.

[57:26] So, yeah, I think there remains some challenges and I think the challenges are less obvious. They can be internal or on small scale. On the wider scale, there are organizations that are supporting women entrepreneurs. There's a lot of push from outside that I think is allowing growth internally. But I think there remains some subtle, not very obvious challenges for women in tech, particularly with, I think, some inherent biases as to their experience and their knowledge and their strategic mindsets. And there still remains some undermining of these skills, potentially because men have done it for longer. But I think women are proving themselves and showing that they have all what it needs to be a successful entrepreneur or a successful leader, high level or executive leader in medtech or also in all of the technology domains.

[58:48] Arseniy Arsentyev: Dr. Mai, it was a brilliant hour talking to you and I want to conclude this episode with the usual section on my show, which is Tomorrow's Medicine. And since you are such close to the top of the FloPatch, the first question in this section is what's the next FloPatch? Can you disclose something that you're working on and what's the next step to bring ultrasound devices into a new level?

[59:18] Dr. Mai Elfarnawany: I think our next phase of focus will be expanding to other use cases. So sepsis, as I mentioned, we have hit 10,000 patients. We've helped 10,000 patients with FloPatch and sepsis today. That continues to grow over time. But I think we are now looking at the other use cases where FloPatch would be very, very helpful and very, very impactful and exploring them currently. And as we explore it, there is a whole lot of things that we do. We do clinical analysis of the fit. We do the tech feasibility. We plan for the future of the development and the tech stack. So it is, yeah, typically takes a couple of years to get something to be very ready and user-facing.

[1:00:07] So we are focusing now on the next use case as we have established some firm grounding on, firm ground on sepsis use case.

[1:00:17] Arseniy Arsentyev: How long do you think it will take? What's the roadmap?

[1:00:20] Dr. Mai Elfarnawany: I would say it's a two-year roadmap to get it to be performant to the level that FloPatch is performant in sepsis today and adapted to the workflow and fitting and really welcomed by the users. So for another use case, I think we are going to build on all of the infrastructure that we have built today. So I think it will be more on the research side, the algorithm side that will require the largest investment to make something that is fit for another use case. But all of the infrastructure of the company and the team, I think we're going to build on what we have today. So it will be easier from that angle and it will depend on the complexity of the use case that we choose to pursue for knowing exactly how.

[1:01:11] But we have given ourselves two years to get to that.

[1:01:14] Arseniy Arsentyev: That's a realistic timeline. And from Flosonics Medical core onto the broader healthcare, health tech and medicine market, let's say in five and 15 years, first of all, what do you think will change and what do you wish to change? I usually joke with my guests and say that in 15 years, I want this apparatus to which I'll just lay down for 10 minutes, wake up and it will show everything that's going on in my body and how I can uptune it, how I can boost it. And what's your vision for five and 15 years ahead?

[1:01:57] Dr. Mai Elfarnawany: That's a very good question. In five years, I think the wearable industry will grow. There's a lot of wearables now. And I think with the help of AI and being able to do high level processing at high speed, there will be a boost in the wearables and where they apply. There's a lot of clinical applications where wearables would be helpful. So I think in five years, we will see a growth in the wearables that are AI based or machine learning AI based.

[1:02:44] Actually, you asked what I wish for too. So what I wish for in five years' time: I wish for us to be at a stage where no one's health is jeopardized due to lack of technology. Because technology is available now and we just need to leverage it. I'll give a small example. Personally, I think in OB in Canada, there is unnecessary suffering. There's unnecessary risk that is caused by the lack of implementation of technology during, delivery, for example. There's still babies that are born still birth because of not looking at them before delivery. I think that should stop. I wish that should stop.

[1:03:45] We have ultrasound machines. We have wearable ultrasound machines. We can know exactly what's happening inside and not just risk it. There's no need to take risks when you can see what's inside. So I wish that in five years time, every place where an imaging technology can be leveraged to save a life, it is leveraged. In 15 years, I think we will have very simple diagnosis. Image-based diagnosis for a whole number of diseases. And hopefully cancer can be detected really early on with AI models, with the predictions. Because in my mind, AI is not an invention.

[1:04:36] It is the accumulation of our knowledge as humanity over all the years. Like everything that is going into AI now is happening because some human contributed to that data, contributed to the categorization, the classification. And so it is really the amazing ability to gather all the brilliant minds, all the skills and all the expertise into a machine and allow it to determine something. So my wish in 15 years is to have systems that can diagnose things with very high reliability and predictability. And we won't be guessing anymore. We won't be trying to figure out what this person is suffering from. It would be more like what you're describing, but not head to toe, more for I'm feeling this. What is it? And there's not a lot of ambiguity and a lot of trial and error.

[1:05:37] I think we will grow into a no-need-for-trial-and-error, no-need-for-unnecessary-risk era of medical technology. And I hope to be part of it.

[1:05:48] Arseniy Arsentyev: You will. That was profound. Thank you very much, Dr. Mai, for such an amazing conversation. It will definitely benefit the listeners. Yeah, thank you very much.

[1:05:58] Dr. Mai Elfarnawany: Thank you very much for having me. It was really, really enjoyable. I enjoyed every bit of it. I hope I was brief as well as informative for your audience. And I look forward to seeing and following your podcast and seeing who else will be on it and benefiting from all your guests.

[1:06:19] Arseniy Arsentyev: Thank you, everyone. You know what to do. Subscribe to Tomorrow's Medicine. Subscribe to Dr. Mai's pages and leave a comment whom they want to see next. Bye bye, everyone.