COMMERCIALIZATION PITFALLS featuring MICHELLE LOTT AND RENAE FRANZ

Michelle, you're muted. Are you gonna promote the people. I'll sign out and come in as me type. There.

I've got to go get her remote control. I'll I'll be right back.

Well, she might have a lot to say during this. How you doing, Sean?

I'm well, how are you? I just.

I'm well.

Saw a message come in from you.

Oh, I wasn't sure like, if you wanted us to log on earlier. I I just haven't done one of these Mdg. Things, so I didn't know.

Oh, yeah, no. I was just doing 800 other things.

Okay.

I'll be right back. One of the other students was making a cup of coffee.

No joke, of.

I don't know. He saw me, and he left. Oh.

John must be running this today.

Yeah, yeah. So I hope we have time to get through everything. If not. we'll just have to do a part 2.

This is a very casual group, and so it the conversation will free flow. So I wouldn't be too focused on trying to get trying trying to get through everything. Everything. Because, okay, I think I think I think organically, we will. Does that make sense go exactly in the order, and then also people share their own stories about like. So you might share your one of your experiences. Then they'll they'll then then they'll like. Oh, that reminds me of the time that such and such. Hmm.

Non co-host. Me.

Where are you at, Joey.

I'm in San Diego today.

Nice nice to see you, Joe. It's been a long time. It has been a very long time, and we're gonna get to see one another in person for the 1st time in May. That's right. In Newport.

Where? Where are you staying in San Diego?

I'm actually at the Grand Hyatt apparently is the.

Oh, I love that place. Yeah.

Hotel in San Diego, and I have access to the club room, and the breakfast itself is like a $40 spread. I had bagels and locks, and they had the whole European.

Excuse me, it does not.

I'm like Joey.

I know, but it was all points, and I came down to pick up Jennifer after a bit of a stay. So it's like a a little holiday for us.

Nice.

So I I went big and.

It's nice to see that you can like. Do it.

Once a while. Yeah, every once in a great, great while. How are you feeling? Are those new glasses? You look nice.

Thank you, Joe. You have asked me that about these glasses for.

Time you wear them.

Yeah. Well, no, for like 3 years. And cause I this is the only para we're on camera. Why is that still sweet? But thank you. Oh, because this is this is our. These are my computer glasses. These are my, see something across the room glasses or watch TV. But I can't like see that they're too much for the computer. Also, these have prisms in them that make my eyes go apart. These have prisms in them that make my eyes come together. So it really messes with your brain. If you're wearing the wrong set of prisms to do the wrong thing.

Yeah, tell me, if you have prism problems on the call anyone else? Is it just Michelle.

No, it looks like Nelly does so.

So I have 3 forms of lazy eye, which means I have.

You're really lazy.

I am very lazy, I know but I have. So I have computer glasses. And my brain knows that when I sit in front of my computer. I'm supposed to have my computer glasses. So if I sit in front of my computer with my regular glasses, my brain like, Nope, it's all got to be fuzzy here. It's really hysterical, it's like, and I have driving glasses, and I have sunglasses, you know. Walking Glasses.

Yep, and I'm wearing minus 12 contacts under beef.

Alright. Do you have your worst? I can't wear contacts because I've had surgery on my eyes, and I have a football shape where the contact would just spin. That's what the guys told me.

Oh!

Okay. Yep.

So I didn't know you used the green screen.

Cool.

Yeah, just to block the sun out. It's so bright behind me.

It seems to be very bright behind Daniel as well, although I strongly suspect that he's not overlooking the Golden Gate Bridge at the moment.

Well, it's pretty bright behind you, too, Joe, but that eliminates the important people.

Daniel, where are you today? You're in Massachusetts.

No, I'm I'm I'm in the Bay area. That's where I'm I'm normally located. The rest of the team is back in Massachusetts.

Hmm! How often do you see the rest of the team? And how often does the rest of the team see their granddaughter.

Yeah, virtually. I see them every day, you know, but he comes out probably every couple of months to see this grandson.

I was like, did you have the other baby.

No, yeah. As soon as I said, daughter, I got got one more on the way. Coming may May first.st

The boy or girl.

Not not early enough to where I can do the circuit at at Joe's event this year. Like I did before.

If you ask me, it sounds like it.

That for next year. Joe.

It sounds like a complete domino. Excuse to me like, Oh, I'm having a child. That's why I can't come.

Okay, we're gonna start. I'm super excited because Thelma and Louise are here today. And This was this was inspired by a webinar. They did a couple weeks ago together, and there was so much content in the thing that we were just rushing through the last. I don't know 15 slides, so we wanted to reprise the thing and bring it back in a slightly less formal manner, but still going through the same material. And it's basically like, you know, how, how do you bring a device to market? And what does it really take? And what could possibly go wrong.

and with that I'm going to turn it over to dumb. And Louise.

Renee and I haven't settled on which one is which yet, though. Yeah, we're we're still trying to figure that out.

Although my grandmother's name was Louise, so I'll take that.

The I don't know, since this is, you know, a smaller, more casual group. I don't know. Do we need to, Michelle? You know everybody right.

Yeah, yeah. You just give yourself a brief introduction.

Yeah. So I've been doing medical device development and commercialization for over 30 years now. And I've helped bring to market over 200 different medical technologies. It says, don't shoot, because I'm always the messenger, because I'm usually the person that has to step in and tell people things they don't want to hear. So that's a bit about my background. When Michelle and I did the masterclass a few weeks ago. This was the agenda that we followed. I know that this meeting is a lot more casual, so I think what we'll do is, you know, we'll just go through the different sections, and then, as you have comments, questions, or we want to have a dialogue, we'll just pause

but we really want to focus on product development. And in your business plan how important it is to have those 2 done right after you have your idea. Of course, regulatory. Then you move into prototyping and manufacturing, and then, like the real race starts. And that's you know, your commercialization and getting that product into the hands of the healthcare providers or the patients. So, Michelle, you.

Oh, yeah, sorry. So you guys are familiar with with this that talk about this all the time that this is where people come to me because they say, hey, Michelle, we've just gotten all the way through our product development and our testing. And we're ready for you to write a submission. And so they they come to us at that point because they think it sounds expensive to engage when they actually need us. But if you look at all the regulatory and innovation commercialization activities and kind of where they should happen all at each step along this way you need inputs from regulatory and your business plan.

because otherwise it really does get expensive. Because you have to get in your time machine and build all of these activities and documentation in retrospect. And that really is expensive. So the problem is that a lot of companies, especially startups that are tech spin outs from universities think that this is where they're at in terms of. They think that they come to us and that they've only got, say, 20% of the activities left to complete. And so they think that they're ready for market launch. But this is the reality is that we see that they have. They think that they have 80% done. But they really have 80% that still needs to be done. And we have failure to launch.

So that's where we're going to talk about the butterfly effect which I'm imagining. Most of you are familiar with. The butterfly flaps its wings on this side of the world, and a tornado occurs on the other. So before we go into some of those stories, one one document that I think is really helpful. And we've we've created it and made it available for you is a trace matrix. So essentially what what it's designed to do in the very beginning. As soon as you have your idea is, get your cross functional team all aligned on what the device could do. So really your your brick and mortar, the the folks that are building the device, the folks that are doing the regulatory, the quality work and clinical. They're really building the foundation.

But that foundation is not going to grow unless you have marketing sales, executive leadership and finance input. And I know you know, most of us work with small companies. So one person might be marketing sales, leadership and finance. But taking that into consideration, you know, I have an example of a device that engineering created through it over the fence to sales and marketing. It was a battery powered lab device, and when sales and marketing got it, and it went into the hands of the customer when the device was fully charged. So people plugged it in at night before they left for work, and they arrived at work the next day, if it was fully charged, all the lights went out.

so people were confused in the lab. They're like, wait a minute. I plugged the device in last night. Shouldn't the lights be on if it's fully charged? Well, then, that was compounded by the issue of there ended up being circuit issues with the device. So then, the lights really were not coming on, charged or not. So that's an example of where it's so important to have the sales and marketing leadership engage with the technical leadership to ensure that the device is actually going to function in the hands of the end user. And we actually had to go back and redesign that circuit. So this is an example of a trace matrix. This is a really high level.

What you will see with many devices, especially the more complex ones. This will be 30 to 40 pages long. So we're just gonna highlight a few of the critical ones where we can share a butterfly effect example with you. But just know, for example, one of these like identify countries, the device will be used in that alone can be 20 to 30 inputs. So the 1st one that we're going to talk about, though, is identifying a temperature range. And this really applies to devices themselves. Of course, not so much software. And AI. So it's really critical to identify, not just the temperature range that the device will be used in.

But the temperature range that it will be stored in. And here's why. So all of the examples I'm going to share are from devices that were launched 25 to 30 years ago. So they're obsolete now, because better technologies have come along. But the examples are still relevant. So in this case we had a power cord, an A/C power cord. We had a purchasing person who thought she was doing a great thing by going out and sourcing a less expensive power cord. Well, we had just launched a whole blood mixing device that was used all throughout Canada that used this power cord, and it was February and the final province to launch was Ontario.

So you know, what could possibly go wrong with changing out a power cord for just Ontario? What happened was we started getting phone calls just from Ontario, and they said, all of our plugs are falling out of the wall. We're like what the heck I mean. It's February, everywhere in Canada. Why is just Ontario having this problem, we flew up there, and the 1st thing I noticed was the power cord was a different color, and it was a different material. What ended up happening is that the new power cord was made out of a plastic. The the devices were being stored on a on vans at negative 40 degrees, and when you would bring the power cord the whole device back in, let's say, in the morning to do a mobile collection at a school or a library.

The cord itself, the part that went into the wall had shrunk because of the cold, so they were all falling out of the walls. So what ended up happening if you think about the butterfly effect? Is our purchasing person was trying to do the right thing and save some money. But the product didn't function properly in the market, and there was actually a blood shortage in the province of Ontario. While we were trying to figure out why their devices weren't working, and it took about a month for us to figure this out and get them back on track, so they had to actually bring blood in from other provinces of Canada. So you know, in hindsight, and you you all may have seen this

before within the work that you do. What should have happened is the purchasing person should have communicated with engineering. Hey? I want to use this less expensive cord. Can you validate both the storage temperature and the operational temperature of it? So then, another really important piece to identify is what countries your device will be used in. And there's a lot of things to think about. So if your device requires a plug in a wall, you need to think about the fact that the Us. Is 1 10 volt. Europe is 2 20 volt. Japan's 100. So there's voltage considerations. There's power plug considerations. You can see here in the lower left hand side. And then there are also cultural considerations like the size of the people in the various countries that you're working with.

Actually, the power is very important. I was in a device at Abbott laboratories 25 years ago. and they went from us to EU. They redesigned the board to accommodate the new power structure. And then for 3 to 5 years we kept having design problems, and what they did. What they realized is they when they redesigned it? They didn't factor in the new powers exposure and the boards were burning out.

You're burning up because you had 2 20 volt coming in instead of 1 10.th Yup. Yep, yeah. So that's a great butterfly effect, Nelly. Another one is, you know, thinking about.

And it took him 3, almost 5, and it ended up.

Am I pausing, or is she? Am I having network.

No, that was me. I'm sorry.

Okay, okay, no. I just wanted to make sure it wasn't me.

No, no, I mean you're that's just one example that I have, which I'm sure you have a many of them as well.

Oh, yeah, yeah. it's a huge consideration for an electrical powered device. Yeah, we had issues, too. I don't know if any of you experienced this, but when Rojas came into play and you had to remove lead, and so we had, you know, all of our circuit boards for this this company. I was a general manager of use, radio frequency, energy. And so all of our circuit boards. You had lead solder and lead in our components. And so another issue that we had. This was kind of a butterfly effect. We we changed all of our components out to be Rojas compliant, and then we launched a new handheld blood sealer and the 1st people to adopt it was Hong Kong, the Hong Kong, Red Cross, and all of the boards burned out. None of them worked, so we had like 50 sealers over there.

and had to go with a team and go change out all the boards, and they were just furious because it held up their blood collection, for, you know, probably 3 weeks, and it was all because nobody thought to test the boards that no longer had lead in the solder. So it was another example. The the other example I wanted to share, too, is again, we're working with this handheld sealer. When we 1st designed it, you can see here at the bottom there's a picture of somebody having a device in their hand. So this particular device you would hold in your hand, and you would actuate it. using a magnet here at the bottom that would at the top emit radio frequency energy that would seal blood bag tubing. And the problem was when we launched the device in Japan, because their hands are so much smaller

they couldn't close the device enough to actuate the magnet so they would get intermittent radio frequency energy at the top, which would spark and burn through the tubing. So then they were losing units of blood just because of the fact that we had a design flaw that didn't accommodate for the size of their hands. So we had to implement what we affectionately called the J. Screw to reduce the throw. On closing the device so we could activate the magnet.

So this us to, you know, regulatory. Not only do you have all these human factors and technical considerations, but then you also have such a variation of regulatory requirements. You know, once upon a time that FDA was known to be difficult to come to market, and everybody wanted to go to Europe first, st but that it's changed, Rene. So now, ironically, FDA is easier to get into, and apparently it's going to get increasingly easier in the coming years, and then everything's still chaos in Europe in terms of expectations.

So. And here's just one example of like FDA. It's it's usually very clear about the standards, performance standards that you have to comply with, and they can be found, like right in your product code that are for the product specific standards. But then it's not quite as clear how many, what kind? And you have to have a whole different set of standards that have to be harmonized within the European Union. Not just. And so you can't use the versions necessarily that were recognized by the FDA. So that's where it's really important to make sure that your trace matrix from the very beginning has all of those regulatory considerations. And then it's not enough I've had in the past marketing people. Come to me, and I'm like, well cut, you know, customers, I'm like, Well, where do you want to sell this? And they're like, you know, globally.

And I'm like, well, if you weren't thinking about every single country or region in the world when you design this, it is not gonna function globally. Also, that sounds expensive. All the registrations globally.

Well, for all the things we've already talked about for power considerations. But also, let's say, if you're developing software with AI and a device. Identifying the countries. Of course it's going to go into is important for language reasons, but it's also important to understand, for example, the Spanish that they speak in Spain is very different than what they might speak in the Commonwealth of Puerto Rico or over in Mexico or South America. So there's there's a lot of really good reasons to spend time initially identifying your markets and outlining everything in this trace matrix. So when your device does hit the market, you have the best shot of having a product that's going to meet that market's needs, and you're not going to have to go back and redesign it, re-engineer it, or have to get it recleared from a regulatory perspective.

And the other question regarding that so is it possible for a device to have a different, a different design based on the market that it is going in. For example, if there's a switch in some countries, it's con considered on, if you flip it upwards. But in some countries it's considered on, if it, you flip it onwards. So you have, like a different design from one country to another country. So is that normal.

That is normal. So you, that's another. That's a really good point. I'm I'm glad that you brought that up, because that can be a factor that you need to consider in device development. And and that leads me to another story which is knowing. Let's say you have a device. That's well, I'm just gonna give you the example. And and it will make sense. And it's based on what you just asked. So we had a blood collection device that they wanted to move from Gram collections to milliliters. So that's a conversion using specific gravity. So we initially designed the device to work here in the Us. And it worked beautifully. We were able to do our blood draws and come in at the accuracy level that would allow the right mixture of blood to anticoagulant.

We went to launch in Japan. I keep picking on Japan, but it's such a different culture than here that we didn't understand as well, and we were overdrawing the blood donors in Japan, which number one makes them sick, and number 2. Then you don't have the right blood to anticoagulant ratio. Well, what we've later discovered is that they have a different specific gravity to their blood. So we're using the wrong calculation. And so it took us a while to figure out we had to pull some clinicians in to help us understand. Okay, in Japan they have this specific gravity of blood, and that's the calculation we need to use in our whole blood collection devices that were sold there.

Yeah, that's a really good example of to just further expand on that. How do you manage that in your Qms, like, just like a very brief overview, because I've yet to quite run into a situation like that. But I want to be aware how to manage it.

So I can't. I can't speak from a quality management system perspective. Michelle could. But what I can tell you is important is when you're asking another great question when you go to this particular part of the user needs in the trace matrix and you're identifying the countries that the device will be used in. Let's say, you know, you know, you know, that you're going to be collecting blood, and you know that you want to move from just a straight weight to, you know, Milliliter collection. You need to have somebody on the team like a clinician thinking about. Okay, I know we're going into a different country they have. We know that the specific gravity of their blood is different. Let's see if that, let's include that in the trace matrix. And let's test to that to see if

there is a different outcome in the performance of the device from a quality management system perspective. Though, Michelle, how would that play in.

And and so you're wanting to see how the quality management system can capture. Like all of these inputs or the complexity of the design.

All of it is like already done for a particular country, but it has to be like slightly different for a different country. So will you.

Oh, I see. So so if okay, so this is another reason why it is important to involve regulatory and quality upfront. Because if you go all the way through the design process, and then you say, Well, now, we want to send it to EU, or we want to send it to Canada. Not only do you have all those technical problems, but you also now have a huge huge gaps in your quality management system. You know, fairly significant, because Canada has some things that are required that the whole rest of the world doesn't have to have. And then, obviously, the EU has a lot of things. And a lot of languages. You know, Canada has a lot of language considerations. So

so so yeah, that that's again, another good point of why you need a re, a true cross cross functional team. Early.

Yeah. And on that point to product cost is a big one. This is where I see a lot of people get into trouble. So for example, I had one product engineering was working on, and they didn't want to be constrained at all by cost. And so, while they were designing my marketing team was out working with customers, and we found out that customers were not willing to pay any more than 2,000 for this particular piece of equipment. but the engineers, the cheapest they could get the design in at was 3,000, and then we had to have a 50% markup because we had distributors. So we had a device that was costing 3,000, that we had to sell for 6,000, that the market wouldn't pay more than 2,000 for.

And so a lot of you know a lot of times, the technical teams will say, I don't wanna have those constraints. But again, having that cross functional team give input like, Okay, that's great. You you don't want those constraints. But then we're not gonna be able to sell the device, or you're gonna have to go back and redesign it in a way that we can position it in the market to move.

The only thing.

Yeah, one of the things I do with my clients when they sit down is like they're all doing all the subcomponents of the parts. And I say, look, let's stop for a second. Let's identify your finished good. So you have a Us. A finished good, and you have an EU or a Canada or Japan, and you go. Now, what is the base part of those finished goods? Bombs that all goes in, all of them? And what are the ones that go in just the other ones. I go break that out. You don't have to have it locked in. but by starting to think that way. Then when you start to build your design requirements and your user requirements, and all of that, you can actually start to factor that in and heard those things in, and then by the time you get to the end, you're just fine tuning it, whereas what happens normally, if somebody goes and builds the device. And now they say, Hey, I want to go to Canada. I want to go to Japan. And now it's like, Wait a minute. Now we've got to build a whole new device

because there's all different requirements. So if you actually sit for a minute in the beginning and kind of map that out where you want to go. Then you can kind of guide things along the way.

Yeah.

Our friend Michelle Mitchell? Asked in the chat about human factors being a part of device design. And Michelle, you are correct. So the examples that Rene gave like, especially about the the Japan and the their, the hands and and some of the other ones that are, you know, country specific, even those are human factor, specific. So in your trace matrix and also your risk management, you should have a whole list of usability questions that you are. You're going through about human factors. and then also know that again, that's going to be different. You're going to have different human factors needs. And then, if you go to the FDA with a human factor, study that was conducted in another country, they're going to be like.

was the study like they, they're gonna have a huge issue with languages. They're going to say, Okay, even if you did this and tested it in English in this other country were those native English speakers, because they may have interpreted your instructions, even though they were in English different than somebody here. So even if you're going to do your study over there. You need to have those considerations and build that thought into your protocols.

And the other comment I wanted to make is, you know, for example, a class one. Gmp exempt device isn't gonna require you to do any sort of, you know, user user needs analysis or you know any of that type of work. But I highly recommend doing it, regardless of your device classification, because it's just good business practice, and it will help you get a device into the market that will actually meet the users needs, and you won't have to go back and redesign. So it's not just even a right. It's a regulatory requirement for some devices. But I recommend doing it for all whether or not it's required.

The other piece that's really important while you're developing the product is also to develop the business plan. So that's, you know, we spend a lot of time looking at like your market size segmenting the market out segmenting the market out is really important, too, because a lot of times that determines what device design you're going to launch first, st because that's gonna determine. Sometimes it's falling into the hands of a different section. Let's say, of a hospital, or even to the different hands of a different type of clinician. So it's important to understand your clinical strategy whether you have to have clinical trials or not to get your device approved, or you have a clinical strategy just in terms of market adoption, where you want people to try the device and and give you feedback. Of course you have to make sure your financial projections work. That's where you know you get into. Oh, the device. They'll only pay 2,000 for it. But we have to sell it for 6,000. Yeah, that's not going to work. So it's important to understand your business dynamics.

and then aligning your regulatory with your reimbursement, too, is key. So if you have a device that you're bringing to market. But let's say there's a chance that you could make certain claims in the like right off the bat and get those cleared that align with the reimbursement code on the other side that you know the device can be paid for, even if it's not a perfect match with where eventually you want to go with the product a lot of times, you know. I'll recommend doing that just to get the product into the market, because there's a path to having it paid for, and for those of you that work on the commercialization side. Like I do, you know, it's nearly impossible to get a device in into, let's say, a hospital unless there's reimbursement associated with it. So that's really key. It really shortens your life cycle to launch and revenue positive.

Of course you have to understand your competitors. If you're de novo, and you really don't have a competitor, it's important to understand at least what people are doing now and how you're better. So you can present cost savings models, for example, to a hospital really have to nail your commercialization strategy, and then your exit strategy. So always kind of start with the exit in mind. What? How are we? Gonna what are we gonna do to either sell this business? Scale it? That's super important, because it drives all these other areas within your business plan. including your funding, which is key.

So, Michelle, do you want to touch on your regulatory strategy?

Sure. So you guys have seen seen me talk about this, you know. Probably dozens of times. We don't have to go into detail. But you know there's so much innuendo in how the regulations are written, and how exemptions are written that you could go anywhere from a walk in the park to like training for an ironman, all depending on your intended use your indications and technology. So then you you can. You can go through these, Renee. We don't have to spend any time on them.

Okay.

Hmm! Oh, so the consequences of not staying in bounds. I think we took out the claim slide here. So we or is the claim side coming.

No, I get.

Yeah. So okay, that that was that was, this is just missing. So for both Rene and I, substantiating claims is huge for different reasons, you know, for for me. I want to make sure that the claims are consistent with both the regulatory strategy and what got cleared. and so consequences of not staying in bounds of your your claims is that this company, you know, just a year ago, had a 5, 10 k. From the nineties that they had made tons of design changes to over the years never submitted any other 510 ks. Well along the way they changed their technology and they changed their indication, and they ended up moving themselves from a class 2 to a class 3, and having to do a Pma. And this was for implantable things. And so FDA had to send out an announcement

from the FDA which normally recalls, or at the the manufacturer level. But this one had to tell. FDA felt like it needed to tell the whole medical community about these hips. And oh, if, by the way, you have one of these. You need to go see your doctor about what what to do, not necessarily to get it explanted. But but you just need a plan. so.

Yeah, and from.

You're calling in from calling in from Pakistan. It's quarter to 10. We better let him ask his question before he has to go to sleep.

No worries. I was wondering, like when you mentioned claims. Is it about marketing claims, or does that also include like claims that are within a patent.

All of the above, because marketing claims often will like get real fuzzy on the regulatory side. Because if you say something like it's twice as fast or 50% better, or preferred. All of that sounds like marketing things, but it all has to be substantiated, and the FDA will call you out on that claim if it either wasn't in your 510 k. Or you cannot substantiate it.

What about the ones in the patent? The claims within a patent.

So that's that's a that's a really good question, too. And I'm glad that you asked that because one of the most common things that I see, and this may, I think this indirectly relates to your question. but it's it's equally as important. So the claims that you make in your patent when you go to launch a device they may not always match the technology that you originally developed. So what I see commonly is somebody started developing a technology, let's say 10 years ago. And it's been through so many design iterations that when we finally get regulatory clearance and we are launching in the market. The claims that they got cleared

do not match how the device functions. And so I recommend. And this is actually part of what we were going to talk about under an IP strategy is that you are continuously reevaluating your intellectual property, and especially the claims that you have on your issued patents.

I'll ask the question on behalf of freshly cash. I invited him to come on camera but his daughter is singing right now he's in Umberg, and I went to Linkedin to look him up. You wouldn't believe how many Hershikesh despandes there are. So he had to connect with me. He wants to ask a question about AI enablement, and how it requirements change depending on the country. I wonder if he recognizes a complex question. How well versed are you guys in? I know it's evolving.

But is it evolving at different rates in different parts of the world?

I mean, you all probably know the FDA just had massive layoffs. the what 2 days ago Michelle and most of it was in software. And AI.

Yup, digital health and AI.

Yeah.

Yep.

This is a.

And then you look at the the current where they're going like, and and the people who are making the decisions they're they're heavily into AI. So I think in the Us. Like we can foresee that it's going to be less regulated even than it currently is.

But I think to answer his question, depending on the market that he wants to go into the country. It's gonna have to do a deep dive on their regulations, because even here it's not clear. So we we have a lot of AI coming into us now and depending on what you're trying to accomplish. With the device, too. It may be regulated by other entities, such as we have a client that has a device that might be regulated by FDA, by NIST, by the American Medical Association, by the American Hospital Association, by the Cures Act. So there's all these different regulations, because AI is so new that may overlay on top of the traditional medical device. Regulatory path, too, that you need to understand. So it's a great question. But it's very complex, and you just need to go country by country.

He's actually doing a great deal of work on machine learning, as it relates to healthcare. I just wrote him a private note. I don't know if coming to California is possible, but I can definitely see his area of expertise adding a lot to our conversation in May.

But Rama asked in the Q. And a. What kind of insight can you offer when choosing the submission path for AI enabled medical devices. And that, again is is really up in the air right now, based off of how many people and the particular areas that they got laid off. You know, typically the even moving to an AI like versus a machine learning model can could possibly change. You know, you're this the submission type. So I don't have a real clear answer to that right now.

And a lot of it. I think Michelle, correct me if I'm wrong, will be device specific, like, what are you trying to accomplish with the AI.

Bye.

You're just going.

I would like to add one more thing on it. There's like a IC 6, 2, 3, 0 4 draft that was announced a few weeks ago. That also includes, some sections regarding cybersecurity. And AI we have to like create an account to see a draft over there. But that might help a bit. It will be released like next year, I believe.

Does anybody here like? I guess we don't have Jeff Gable or anybody here that does primarily only software that can speak to possibly speak, speak to that.

You do go ahead and hit the raise your hand button on on the panel. Rama, do you feel that the ladies addressed your comment adequately. Do you have a follow up.

I mean, I think it's just difficult right now with the what's happening with FDA and just choosing when you have these small changes to a standalone Samd product. And you're not sure whether or not this is a significant change that warrants a new 510 K just based on like minimal impact to maybe intended use. Or you know the validations that are related. I think there's a lot of gray area. Still, when choosing whether or not, you should do a letter to file, or a Pccp. Or a new 510 KI think I think it's just a little bit difficult to answer. So that's why, you know, it's it's great to get these questions out there.

Yeah. It's difficult for everybody right now. It's.

I've shared the link to this draft of the standard that might help you a bit regarding that.

Okay.

You.

I was kind of wondering because AI is so new, where you thinking it's gonna fall in a classification, or has that even, I mean, because everybody's saying, Do AI do? AI, and it's like. Well, are there going to be different classifications? Is it gonna be the same classification? Is it going to be a new classification, because to me that seems a lot more complex to from a design requirements thought about it.

You. You might have heard me advocate for pre-submissions more than one or 2 times.

Yes, I have.

So that though these are the exact type of conversations that you need to have with the FDA, because your understanding of your technology and their understanding of your technology. And all those innuendos are are different. So I think that that in that area where you're especially kind of on on the cusp of both things that they are regulating and things that they don't have really good guidelines or understandings. For is you need to have those probably even multiple pre-submissions as you go.

And Michelle, wouldn't you also say you need to go in and be very clear with them about your like, your plan and what you think the strategy is, and maybe what other organizations may also be guiding the AI development.

Yeah. Well, the challenge of it is, is I under? I totally agree with the precepts as many as possible. I actually had a client that they're 5 years in getting ready. They were going to do their FDA submission. They were. They had multiple precepts. They thought they were on the same page, and when they went to do the last preset before they started to put their 5, 10 K. Package together, they changed, and all of a sudden wanted clinical trials of like 500 people and like, Wait, wait! We've been talking to you for 5 years. What's what changed? So that that is, that is a potential. And then they did go through the FDA advocacy, and they're actually getting that rolled back. But it it is something that you have to be kind of aware of that. You're you can have your precepts, but it isn't 100 guaranteed.

You actually scooped me. I didn't realize that those job cuts were announced this week. Was it targeting just those 2 groups software. And AI is that where the layoffs took place.

It was anybody in the probationary period. They laid off everybody. So even if they went out of probationary this week, they were laid off. So I had a lot of friends that were.

I'm sorry. I don't know what that means. Do you mean like they were new hires.

Pretty much so if you move to a new group, you're put in probationary for 2 years. So anybody that was within that group was laid off. they received a notice.

Let me paraphrase anyone who either joined FDA or moved.

Moved.

In the last 2 years was laid off.

Correct, correct. So I have a lot of friends from Johns Hopkins, who, like moved from this division to digital health, let's say, or from digital health to ent or or vice anywhere within the organization or from. you know, this division of Nih to a different division. So you're put in a probationary period for 2 years. and so within that if you had the words probationary, you were let go.

You know how many jobs that affected.

Wraps published an article so like, I know the dental division last 18 digital health. I think this. What did they say? Michelle, 8, 89, I think.

Wipes out the whole thing I mean.

Over 200 in digital health.

In digital. It was significant. But I do know, like within, I can talk about artificial intelligence within the cures act. They refined the oversight within digital health. And then I they just published actually a guidance within digital health. And I was trying to find that link. I just had it, and so I apologize for not having it right at my fingertips.

Digital Health Group 2 years ago was so small that, like there can be nothing but a skeleton crew left. If the definition right let go was any.

Yeah, with, patel.

Years.

When he left. Then it was, and then they had a workshop on the digital health and 25, and they just released a thing. And most of it's around the G Gmpl setting, and it's set for the best practices in the aiml. Have you all seen that circular artificial intelligence. It's from the Institute of Electrical and Electronic Engineers at p. 28 0. 1 working group Subcommittee on Artificial Intelligence Forum. When I find it, I'll put it in the chat, so I didn't mean to interrupt your.

No, no, that's great, Karen.

Well, I know you're.

Camera shy today. But if you want to.

Oh, did I go on? Oh, I'm not I! When I did my.

Do you want to just say aloud what you.

No, as you guys are talking, I just googled it. And there's this thing that I posted in the chat about the number of

Oh, the employees!

Total, like.

Oh, yeah, it was disgusting.

And have been just 200.

And.

But you know, if you think about it, this could go one of 2 ways from regulatory perspective, because, you know, during and after Covid. They had to hire so much it it, and it has taken them almost 3 years to catch up on just the pre sub term process because they got 3,000 pre subs in like 2021 and and that was like far blew away anything that they had resources for. So they were bogged down for a long time after Covid was over, with just the number of submissions in the queue. So there's a possibility that exists that now we're far enough past Covid that they are. Have that they are too heavy on their headcount. And now it's time to like kind of pull that back down to you know

what it was kind of Pre Covid.

We speculate that the people who.

That there'll be a reshuffling, and that people who have tenure will backfill.

I hope they have tenure again. They sure were green. That's but the ones that they did hire. In my opinion they were experienced in a lot of those. They got them all back, but.

Okay. Well.

I know which is unfortunate. A lot of the ones they did let go were some of the older ones that came back to the agency, in my opinion, but I'm trying to find this new guy AI guidance that they just had. I had it. I'll add it in the chat, but this is the old went in there, go ahead with your.

Yeah, I was. Gonna say, given that we've kind of gone into something quite

Yeah, well, but it's timely considering everything you know what's going on. And the fact that there's so many device, you know. technologies coming that have AI and software. this next section is about basically actual devices themselves. They may have software and AI, but these are like, basically, you know, devices that you would manufacture. So you know, a lot of times we recommend that folks 3D. Print a design just to get start getting feedback on. Okay with key opinion leaders. Is this something that would work for you in the field like having clinicians work with it, or even the engineering teams to do some initial just verification testing. Does this design that we created meet the needs?

Well, 3D. Printing is wonderful for that. But as Michelle will talk about here in a minute. It is not something that 99% of the time you could actually submit to the FDA because you're using a totally different printing method. In most cases you're going to go from 3D. Printing a device to some form of injection molding. And so you're going to have different materials that come out of injection molding. They're going to have different properties. And so what ends up happening that I see a lot is you'll have a beautiful design that comes out 3D. Printed, and then you have to do a redesign for the injection molding process because of the way the materials flow through the mold and the whole molding process itself.

So other things to keep in mind. Let's say you know, your device is going to go through biocompatibility testing or needs to be sterilized again. What you've just printed with your 3D prototype is going to be a different material than let's say what you're going to injection mold, so you can't use that prototype. And you also need to take into consideration. Let's say your device is going to be sterilized, and you're using injection molded parts. You're gonna have to mold those parts, sterilize them, validate your sterilization process and then run your biocompatibility because there could be structural or chemical changes to the device through sterilization that altered the results of biocompatibility.

The other thing that I see a lot with with clients, too, is they'll 3D. They'll design a device, it'll, you know, look great work, great 3D. Printed. But when you go to actually manufacture it, a lot of the walls aren't thick enough, so you have to do in some cases modifications or even design changes to what you've designed coming out of 3D. Printing in order to be able to injection, mold it and have a device that's going to be functional in the field. And so a lot of times. What I'll tell folks is, you know, beautiful prototype doesn't necessarily mean that you can manufacture that.

Have you done any? Have you done any 3D. Printing at your shop?

Yeah, we do it all the time. We do several iterations for our customers. I'll get the look and feel, and so forth. And you know, Rene's 100% correct, but then we have to change draft angles and so forth for the injection molded parts. Sometimes we'll use Urethane casts because those might be a little bit closer to the injection molds. And we've had. We have presented to the FDA some Urethane cast materials. But then, you know we do have to go back with the injection molded parts because they are different.

Could I think of 3D. Printing, as it relates to devices as strictly prototyping stuff? Or can you scale with 3D. Print.

You can scale. There are certain components that you can use. You can't really use it for, let's say, an overall overall enclosure, but you can use it for certain parts, and so forth, that that can be 3D. Printed.

And it would make sense economically to have a part of the device be mass produced in 3D. Printing.

It could be if the if the part is complex, to do an injection molding, if there's some undercuts and so forth, then sometimes it's 3D. Printed is better than an injection mold.

Thank you.

So this is one of my favorites. Is that you know, a startup, especially startups will often like try to pursue a manufacturing partner based off of price or somebody, or a mom and pop shop that they know and either the contract manufacturer doesn't have a quality management system or and is not willing to build one or register with the FDA. Once they realize it's $9,200 and more a year. And so then the the startup is kind of left holding the bag, and because they don't have a product that's manufacturable because the contract manufacturer can't legally manufacture it, or register with FDA.

In fact, Michelle, we have customers that, let's say, are a specification developer, and they don't have any quality management system at all. And they hours, you know, because we're we're FDA and Iso 1345.

Well, yeah. And and like, even if they outsource the bulk of the quality management system to somebody like like you, that's highly competent. And that doesn't mean that that that they don't have to have any procedures, because, like they have to qualify you as a supplier to start with, they should be having some sort of management review. And like, you know, there's a basic set of procedures that no matter what you can't outsource. And so you're right, Andre. There's problems on both sides of that that coin. So this, this is mine. This was the 1st recall I've ever done in my career it. It was a resuscitation bag, and the part that that the l bracket where it holds the mask on was that whole thing was like falling apart while they were trying to resuscitate somebody.

So if we looked at the root cause of what was happening, you know. Talk about not designed for manufacturing. It wasn't designed for manufacturing, it wasn't designed for for assembly. It had problems with flash on a mating surface. The print specifically said flash is allowed there. And it's it's a making surface like why would you say that? Well, they had a problem with their validation, and they couldn't get it hold in to make it without flash. They're like, well. let's change the acceptance criteria. So that was at the molded component level. And so then we took all of that. That was a validation lot that didn't meet it. We put it in assembly where they were supposed to be poki oaks, to verify that that the press fit was like solid. They weren't using it because it was on like it wasn't in their process flow. And the assembly process was way too complex for something that should have been press fit.

And then last, but not least, the what nearly sent me to the crazy house was that the er Sis. The Erp system did not like it would allow you to build a car and not put spark plugs in it because it didn't have spark plugs on the bill of materials in the Erp system to pull them from inventory. So then, I had to physically go through every single lot of these products, for, like the for 3 or 4 months, to find every one of these manually, because all the letters I sent out were not about the recall were not complete with all the lot numbers.

Yeah, I just had instance, Michelle. I I left the company, but they are the regulatory person called me because you won't believe this. So when they were doing Bnb. so they have a device that has small medium, large sizes. Okay? And when they were doing Bnd they were only ordering a small size or and then, when they went to commercialization, they added a second size to their lot. and then they added a 3rd one. Right well, they placed the order. and because all the sizes were ran on the same line for different fgs's of different sizes, they put the same lot number on it all 3. And so nobody knew that in the beginning. So that's the other thing I tell people is, look, figure out your finished goods and your lot numbering conventions because you can't guarantee that if they're Iso 9,000 or Iso 1345 certified

that they aren't having issues like this. And they're like we've gotten away with it. FDA and Iceland have never noticed before. No end.

So got away with it. Isn't that nice.

And so.

I mean, how.

That for a culture of quality right now.

Matter till it does.

Yeah, everything that can go wrong will go wrong if you're not looking for everything.

Bingo, yeah, yeah.

Do you have a do you have another meeting right now?

I have another meeting, and we never we sean we. We didn't make it through commercialization. Yet again.

Michelle. If you can stay on, we don't have to stop at the top of the hour and focus 12.

Almost the whole rest of this is Renee.

It's me. The rest of it's me.

Probably spend a whole other hour some other time on just that portion of it.

That? How's next week?

Yeah, I think we should.

So, Renee, you can go, and I can finish finish up here.

That sounds good, really nice to meet all of you. And.

To meet you in person. Anyone who's coming.

Yep.

And will you? Do you know the title of your talk in May?

Sean and I are working on it.

Okay. Fair enough.

So I'm gonna interrupt. One cool thing that they did at the end of their masterclass is they launched a new program. It's a new marketing program, and it's called Lean Raqa Launchpad. And it is essentially consulting as a service that you can pay for by the month. and you get a certain number of hours and you get access to the full commercialization, regulatory biocompatibility, all of the experts that work with lean. It's ideal for startups because it allows them to get the support that they need in a variety of areas. Advice about. you know this, that, or the other thing, and it is affordable and predictable, and so I.

That's the question. Do you have a fixed price.

Yes, it's 7, 99 a month, and you get a certain number of hours that you could say, Hey, Michelle, have a bio compatibility question, and I'll introduce you to my.

Less than $800 a month.

Yes.

Wow, thanks surprisingly inexpensive. How many can I get.

Gateway drug. It's a gateway drug. So you know.

Okay.

Be, be, be be forewarned, you know the 1st time is for.

Once you get a little Michelle, you're like.

Yeah, that.

7, 99 thing. I want the 7,999, one.

But you know, the thing is is that we designed the program because it's very difficult to wrap your mind around the idea of oh, do you want to hire a consultant? No, I don't. It's like, Do you want to hire a marketing consultant? No, I don't right. But when you put a package on the thing, and you put it in a format that people are are used to. Then they're more comfortable with it, and if you de-risk the offer like you can cancel with 30 days notice, you know. Then then suddenly it becomes a lot more appealing. So the whole goal.

So yeah, I think you were getting there, Sean. But so if you think about that timeline slide I put up at the beginning. This was, we put this program together in hopes of trying to get people earlier on that timeline because it was less intimidating because you know how much you're gonna spend, and then you would at least ask questions so that you could be informed and not enough to not make such big and costly mistakes as I see people that come to me too late in the process.

I think there's some marketing copy in there somewhere, Sean.

Hmm.

Do we want to do top line commercialization stuff, or shall we just shelve it for part 2.

I think she'll vote for Part 2.

Alright. Well, I'll put you in charge in charge of figuring out when Part 2 is.

Lot of people would be interested in in that piece.

And and is there any consideration to cancelling your other event and vacation so that you can come to 10 x instead? Is that something you and counter talking about.

That? Well, I mean.

That'd be a part.

It's gonna be. It's gonna be. It's gonna be up to me if I can, if I feel like I can physically do that for, and then, like the very next week, be in San Diego, and then the very next week, I gotta audit in Denver that they want me in person. So.

Maybe if there's enough love from the group like a ground swell, like a wave kind of thing that.

I may. I don't.

I mean come all the way in from Pakistan, so I think you can. You can do it. I'm just pulling it from Quebec, and who knows where Michelle is now.

Hey, Michelle, if I could just add a quick note as you guys were talking, I just kept seeing. And, by the way, I'm not involved with extra ink anymore. I mean, I can could be not an active client. But there's just for the sake of any medical device company. There's such a need for supply chain considerations. You know, when we're talking about the iteration, when you were talking about a lot numbers with all those, and I don't know. It'll just be a nice conversation to have to bring the need for that level of insight and consideration along everything that you're doing with your clients.

We haven't had a good supply chain conversation in a while. You should get back on our calendar, Paul.

Okay, yeah, I can. Now, yeah, things moved. And I'm a free agent again.

We also had a conversation recently with somebody who is a Us. Distributor for foreign manufacturers, and does warehousing. Was anybody interested in hearing about that or not.

Yeah.

Yes, yes.

I'm involved landed recently. Sean, are you talking about land and lack.

No.

No, okay. He's in Japan.

You might have to go awesome conversation. See, you guys, next week.

Okay.

Good night.

I'll let you wrap up if you. if that's what we're gonna do, we're gonna follow up with Michelle in another week.

Carrie. Thanks for the contributions today. That was valuable. Yep.

Michelle Mitchell. You wanted to.

Yeah, I just want to say to Michelle, thank you. And I really love your website. And then, I already register for 10 X. But I I might not be, might not able to go. But if you go I will come back from Japan. then go back to Taipei, because I have 2 trips, so if you're not coming, I'm not coming back from Japan and then go back to Taipei. So.

Okay, that's a major commitment.

That's love.

Yeah, I'm gonna I guess I'm gonna have to go. I can't turn that down.

Okay.

If you'll commit to that, I can come back from Greece and go to San Diego.

I know it's like 1030 in the in the evening, but you can start looking up airfare for me.

Okay.

I don't know how much it costs to go.

From Pakistan to California and back.

Oh, my God!

Yeah.

Really like around 2 to $4,000. I I guess something like that.

Well, maybe talk to the authorities and tell them why you're coming, and they'll give you a break or something like that. Bring us home.

Well, I just like to thank Renee, who is no longer here, and Michelle Thelma Louise, for you know, an exciting topic and bringing people out of the woodwork to come to this event. I have so much fun working with them, and we get to do all sorts of amazing things. Lean launchpad is the newest. And I think you're gonna you're gonna see more of this within the industry. This sort of consulting as a service. Arrangements.

Nice before we go, Carrie. I think you should come off. Blur and introduce us to your pet, who wants to be part of this conversation.

Loki. Her name is Loki. It's a good name for a Norwegian elkhound that's a menace.

We get to get to see her or no.

Well, she's now moved on to my child because we have a snow day, because we got 20 inches and.

Jeez.

Denver.

Where are you?

Golden Colorado.

Wow!

Going to be about 70 degrees tomorrow night, and have we have, or tomorrow? And we have baseball. So there we go.

There you go!

Welcome to Colorado.

Good luck with all of that.

Yeah.

Hey? We'll see you all next week. Thank you, Michelle, and when you come, be sure to wear those new glasses.

Bye.

You also come out. No, I can't wear these, because I need to look at the screen. So I have, or the you know the per whatever.

He doesn't listen. Michelle.

I listen. I just don't retain.

He's a New Yorker.

You're my favorite goldfish, Joe.

Right.

This is.

Did it go.

Okay.

Or I get in trouble. That's true.

Okay, thank you. Everybody make the comment.

Thanks everyone. Goodbye.

Bye.