Listen to Audio
Listen Anywhere
March 2026
In this episode, Susan Szathmary and Richard Jaenisch, both of Open BioPharma Research and Training Institute, join the podcast to share how to accelerate the adoption of new technologies through applied AI in pharma manufacturing and for workforce readiness.
1
00:00:00,080 --> 00:00:10,080
Welcome to the ISPE podcast, shaping the future of pharma, where ISPE supports you on your journey, fueling innovation, sharing insights, thought
2
00:00:10,080 --> 00:00:14,240
leadership, and empowering a global community to reimagine what's possible.
3
00:00:15,414 --> 00:00:20,135
Welcome to the ISPE podcast, shaping the future of pharma.
4
00:00:20,535 --> 00:00:22,375
I'm Bob Chew, your host.
5
00:00:22,454 --> 00:00:32,609
And today, we have another episode where we'll be sharing the latest insights and thought leadership on manufacturing, technology, supply chains and regulatory
6
00:00:32,609 --> 00:00:36,049
trends impacting the pharmaceutical industry.
7
00:00:36,609 --> 00:00:43,754
You will hear directly from the innovators, experts and professionals driving progress and shaping the future.
8
00:00:44,234 --> 00:00:45,914
Thank you again for joining us.
9
00:00:46,155 --> 00:00:48,715
And now let's dive into this episode.
10
00:00:49,354 --> 00:00:58,679
Our topic today is the use of applied AI in pharma manufacturing in general and workforce readiness in particular.
11
00:00:59,159 --> 00:01:09,239
To share more about this topic, I would like to welcome Susan Szathmary and Richard Jaenisch, both of Open Biopharma Research and Training Institute.
12
00:01:09,799 --> 00:01:13,224
Welcome to this podcast, shaping the future of pharma.
13
00:01:13,784 --> 00:01:14,825
Thank you for having us.
14
00:01:14,825 --> 00:01:16,105
Thank you for having us.
15
00:01:17,465 --> 00:01:22,745
First, please tell me about Open Biopharma Research and Training Institute.
16
00:01:22,745 --> 00:01:26,905
How did it get started, and what is its scope of services today?
17
00:01:28,740 --> 00:01:32,819
Open Biopharma started in in 2020.
18
00:01:33,299 --> 00:01:34,740
It's a brand new institute.
19
00:01:34,740 --> 00:01:42,100
We have about 45,000 square feet of facility with 6,000 square foot of open clean room space.
20
00:01:42,605 --> 00:01:52,924
And the idea was to provide training in on project in on the job training, focusing on human error reduction
21
00:01:53,325 --> 00:01:54,844
and focusing on The U.
22
00:01:54,844 --> 00:01:55,084
S.
23
00:01:55,084 --> 00:01:55,564
Technologies.
24
00:01:56,020 --> 00:02:02,819
So students and employees looking for new job or riskier or upskill, they can find training on The U.
25
00:02:02,819 --> 00:02:03,140
S.
26
00:02:03,140 --> 00:02:03,780
Technologies.
27
00:02:03,780 --> 00:02:07,459
So this is how we incorporated AI also into our training.
28
00:02:07,459 --> 00:02:10,099
All of our faculty are all coming from the industry.
29
00:02:10,099 --> 00:02:12,675
So they worked at CDMO's pharma companies.
30
00:02:13,155 --> 00:02:17,955
And we also have a great mentor group who are working in pharma.
31
00:02:18,275 --> 00:02:23,635
So one of our major focus is advanced technologies too and advanced therapies.
32
00:02:23,635 --> 00:02:28,050
So we focus on cell therapy and cell and gene therapy also.
33
00:02:28,689 --> 00:02:32,689
And that's basically the core of our things.
34
00:02:32,689 --> 00:02:35,810
It's more like a residency type of training.
35
00:02:35,810 --> 00:02:42,370
We're not the school, but more like the next step, like a residency to get somebody ready for critical manufacturing steps.
36
00:02:42,504 --> 00:02:45,465
We like to do that bridge between, you know, academia and industries.
37
00:02:45,465 --> 00:02:46,985
And there's often a gap between them.
38
00:02:46,985 --> 00:02:56,585
We wanna make sure that, you know, there are folks who are coming out of the coming out of these academic institutions who are right ready to work, ready to and and not just ready to work, like, ready to
39
00:02:56,585 --> 00:02:57,705
work at a at a certain level.
40
00:02:58,349 --> 00:03:06,110
And so, you know, one of the goals of our of our nonprofit, you because Open Biopharma Research and Training Institute is a nonprofit training institute.
41
00:03:06,430 --> 00:03:10,030
And we also have, you know, we have a we have a core lab services.
42
00:03:10,030 --> 00:03:17,405
We have our, you know, our polypractical production space, and we have our our training services that we we offer, and we also host as a venue as well.
43
00:03:17,405 --> 00:03:26,365
So we have often a host, leading training services and training come training from different organizations who are have industry ready training.
44
00:03:26,580 --> 00:03:34,740
And though we often bring do we often work deals with them to have a a channel for all of our, apprentices and interns to be able to access that training.
45
00:03:34,740 --> 00:03:38,180
So this way, then when they come out, they're that's how we say when they're industry ready.
46
00:03:38,180 --> 00:03:47,215
They're they're able to do all of these other things and actually try out these different elements, work on the core lab services, and really get a real well rounded, level of experience to understand where
47
00:03:47,215 --> 00:03:48,735
they best fit.
48
00:03:48,735 --> 00:03:52,895
Because oftentimes, it can be difficult when you're in college trying to figure out exactly what you wanna do.
49
00:03:52,895 --> 00:03:56,334
And then that never really matches with whatever industry is about to do because it always changes.
50
00:03:56,400 --> 00:03:59,919
You know, two years, five years, that's difficult for these things to work.
51
00:03:59,919 --> 00:04:02,080
So that's Open Biopharma in a nutshell.
52
00:04:02,719 --> 00:04:08,000
And, yeah, we're we're happy to be here, and it's a great it's a great organization.
53
00:04:08,000 --> 00:04:08,479
Thank you.
54
00:04:08,479 --> 00:04:15,974
So I've visited your facilities, which are located, what, in Carlsbad, North Of San Diego.
55
00:04:16,454 --> 00:04:27,220
Maybe tell our audience a little bit more about what you've got in the labs and in the development or production spaces?
56
00:04:28,740 --> 00:04:37,584
So we have clean room space, and we're focused on having equipment like single use bioreactors, chromatography systems.
57
00:04:37,584 --> 00:04:47,584
So the very basic ones, everybody can train on on the basic elements, can train on how to connect them, how to gone, how to do environment monitoring in a
58
00:04:48,029 --> 00:04:49,470
in a GMP setting.
59
00:04:49,470 --> 00:05:00,875
And then we also have a lot of quality control type of equipment in our facilities, so flow cytometers, the absorptors, sequencer, PCR
60
00:05:00,875 --> 00:05:03,754
machines, you know, chromatography.
61
00:05:03,754 --> 00:05:11,754
And then we're also working on new technologies, for example, for cell and gene therapy, how we can liquid biopsy, their culture Yeah.
62
00:05:12,089 --> 00:05:21,290
During production and get more information what's going in the cells so we can characterize EVs that coming out of cells and live cells and work on those.
63
00:05:21,290 --> 00:05:27,685
And the students are basically exposed to all the technologies, and we have a rotation.
64
00:05:27,685 --> 00:05:34,165
Make sure they 100% pipette, 100% can work in a septic environment.
65
00:05:34,564 --> 00:05:38,805
And then they familiar with upstream, downstream and the quality control technologies.
66
00:05:39,300 --> 00:05:39,699
Yes.
67
00:05:39,699 --> 00:05:42,019
And we have a number of different locations we do that.
68
00:05:42,019 --> 00:05:45,300
We have a training we have a training lab specifically dedicated to this space.
69
00:05:45,379 --> 00:05:49,939
We have our upstairs spaces, which are more, you know, your basic training rooms, more for our venue operations.
70
00:05:49,939 --> 00:05:57,074
But we also have a dry lab and wet lab upstairs as well for folks who wanna be able to have kind of a mixed mode experience depending upon their training.
71
00:05:58,275 --> 00:05:59,475
Well, that sounds Sorry.
72
00:06:00,275 --> 00:06:02,915
That sounds very comprehensive.
73
00:06:04,899 --> 00:06:12,500
But tell tell me how your use of AI, is improving how you are training your students.
74
00:06:13,699 --> 00:06:23,384
So when it comes to AI, we started working with it very kind of oh, and not really early on per se because, you know, AI has been around since nineteen forties, technically speaking.
75
00:06:23,384 --> 00:06:23,865
Right?
76
00:06:24,345 --> 00:06:30,985
So but what we're talking about when we use AI is, generally speaking, we're using, we started using generative AI.
77
00:06:31,305 --> 00:06:40,580
So what that means in our case is that in 2000 in 2023, all of our we started with our apprentices, and we went, okay.
78
00:06:40,580 --> 00:06:46,420
How do we make sure to give them the readiness they need to work in the in the in the future for the future workforce.
79
00:06:46,420 --> 00:06:46,900
Right?
80
00:06:46,900 --> 00:06:50,154
And so what we did was we started identifying.
81
00:06:50,154 --> 00:06:57,995
We saw there was a there was a paper from the federal from the Fed that explained what the strong suits and weeks and weaknesses of are these are these tools.
82
00:06:57,995 --> 00:06:58,634
And we said, okay.
83
00:06:58,634 --> 00:07:01,675
Well, it seems we're weak research is a bit of a a weak point.
84
00:07:01,675 --> 00:07:01,914
Why?
85
00:07:02,310 --> 00:07:09,430
And we knew that research is a very critical element, these students are learning, and these apprentices are learning how to really conduct appropriate research.
86
00:07:09,589 --> 00:07:11,430
And so we wanted to really test that.
87
00:07:11,430 --> 00:07:21,245
So we started doing what's called a parallel project where everything that they would do, they would have the they would train what, what we'll call an AI assistant to do at the same time.
88
00:07:21,245 --> 00:07:24,845
And so they would iterate this forth back and forth, and they've been doing this for years at this point.
89
00:07:24,845 --> 00:07:27,004
So every cohort that comes on does the same thing.
90
00:07:27,379 --> 00:07:32,019
And so everything that they're doing, when they do it, they do it in parallel with the actual AI assistant.
91
00:07:32,099 --> 00:07:33,220
They create the assistant.
92
00:07:33,220 --> 00:07:34,339
They build the elements in there.
93
00:07:34,339 --> 00:07:45,264
And so that this way, what they're doing, creates a little bit of that that tacit record kind of gets a little bit more under better understood because it also means that when, inevitably,
94
00:07:45,504 --> 00:07:53,985
the apprentice or intern leaves because we are an institute that takes an apprentices and and and interns on in six month, year, two year, they differ.
95
00:07:54,144 --> 00:07:56,625
And so a project, you know, might suffer from a gap.
96
00:07:56,625 --> 00:07:59,839
And so this allows the next person to kind of pop up and go, oh, look.
97
00:07:59,839 --> 00:08:00,560
Here's the assistant.
98
00:08:00,560 --> 00:08:02,240
Let's let's help them kinda carry along.
99
00:08:02,240 --> 00:08:03,919
It ends up kind of carrying these gaps forward.
100
00:08:03,919 --> 00:08:09,839
And so it ended up being a really useful tool to kind of carry along this tribal knowledge that normally kind of can get a little bit lost along the way.
101
00:08:10,319 --> 00:08:18,295
It also meant that in this way, they were also testing their own critical thinking skills because that was one of the elements that I found out early on is that it was bad.
102
00:08:18,694 --> 00:08:28,879
Early generative AI in in in 2023, was, you know, maybe at best, had hit about a 10% ratio in terms of the success rate to actually
103
00:08:28,879 --> 00:08:31,759
get the link and the name and all the rest of materials right.
104
00:08:32,000 --> 00:08:33,679
Now maybe it's about 40%.
105
00:08:33,679 --> 00:08:34,399
It's getting better.
106
00:08:34,399 --> 00:08:35,840
And this is off the shelf tool.
107
00:08:35,840 --> 00:08:38,320
So when you really hone something, and that's what they learned.
108
00:08:38,320 --> 00:08:40,480
When they really honed something, they managed to make it better.
109
00:08:40,865 --> 00:08:42,625
But we're not just using it there.
110
00:08:42,784 --> 00:08:48,625
So that's what we started working with it there to make sure that all of our staff have a really good understanding as well as all of our apprentices.
111
00:08:48,625 --> 00:08:57,550
Anyone who really walks through our doors, we wanna make sure that they're a little bit better off, you know, understanding what the AI landscape looks like and how to integrate it into their workflows.
112
00:08:57,710 --> 00:09:05,470
When we think about the other enhancements that we we're solely putting into play is we have a lot of sensors and cameras, a lot of data that we've been collecting.
113
00:09:05,470 --> 00:09:09,309
We're trying to find new ways of leveraging that, and we'll get into that a little bit later on.
114
00:09:09,309 --> 00:09:19,465
But where it happens with these with these apprentices is that they get to see actively, you know, how this change occurs over time with their assistant
115
00:09:19,465 --> 00:09:22,345
and see how their assistant is kind of, like, almost a partner to them.
116
00:09:24,059 --> 00:09:25,740
Well, that sounds very interesting.
117
00:09:26,299 --> 00:09:33,259
So besides working with your students, how are you using AI today for your day to day operations?
118
00:09:35,125 --> 00:09:41,845
So one of the things that we've been doing is that we have been, kind of upgrading our sensors and upgrading our data acquisition.
119
00:09:42,085 --> 00:09:52,110
So in order to properly install really generative AI, AI, predictive AI, whatever AI you wanna put in there, you really need to have good, not only good sorting
120
00:09:52,110 --> 00:10:01,070
of data in terms of data governance, you need to have a plan as to what sensors you're plugging in, how you want them to work into your workflow, how do these things make sense, because if you don't have
121
00:10:01,070 --> 00:10:01,950
that collection that's there.
122
00:10:01,950 --> 00:10:04,670
So that's really where a large portion of where the steps are.
123
00:10:04,754 --> 00:10:14,835
We're also while we're doing this, we are finishing up the pilot of our of our initial element here, which is a, an integration where
124
00:10:14,835 --> 00:10:21,669
we have a a program that has generative AI elements into it, but it isn't fully generative AI.
125
00:10:21,669 --> 00:10:24,470
It's a combination of a few different types of AI.
126
00:10:25,350 --> 00:10:31,029
My my brother is the architect here for this, so I have to kinda lean on him for the specifics of the engineering behind it.
127
00:10:31,429 --> 00:10:33,190
I'm just the guy who communicates the elements.
128
00:10:33,190 --> 00:10:42,365
So when it comes to it, what it ends up doing is it allows us to access our inventory, access our systems much cleaner, much easier, and importantly, in a way that's accessible.
129
00:10:42,524 --> 00:10:51,090
So one of the biggest changes that we have regarding our the tools that we have built for our own internal management, so inventory and other elements.
130
00:10:51,090 --> 00:10:53,410
It's almost like an ERP system, but it's not quite there.
131
00:10:53,730 --> 00:10:57,649
The idea here is that we we look at it and we say, how do these things connect?
132
00:10:57,649 --> 00:10:59,009
How does this work together?
133
00:10:59,570 --> 00:11:06,324
And how do we how do we make sure that we're best leveraging all of our materials and all of our time?
134
00:11:06,324 --> 00:11:11,365
It's kind of like a little bit of a resource management, a little bit of better resource tools because, you know, we're a nonprofit.
135
00:11:11,365 --> 00:11:16,084
We have to leverage with what we can and what we have, and we're we're scrappy at the end of the day.
136
00:11:17,320 --> 00:11:22,120
We'll scrapping a 45,000 square foot facility, as much as that happens.
137
00:11:22,519 --> 00:11:32,675
But, the the main thing about this, though, is that this allows us to because of the generative AI element integrated into it and because we've worked with so
138
00:11:32,675 --> 00:11:42,434
many of these, not just apprentices, but also our our, our various, employees, myself, I am a a person with a few different disabilities if my glasses don't give me away.
139
00:11:43,394 --> 00:11:49,970
And and what that means is that it can be difficult for me to read certain types of screens.
140
00:11:50,210 --> 00:11:53,009
Like, for instance, I have visual dyslexia.
141
00:11:53,009 --> 00:11:55,730
When I look at a screen, I I kinda mix up letters and words a little bit.
142
00:11:55,730 --> 00:11:57,889
Actually, I mix up concepts, which is a little more complex.
143
00:11:57,889 --> 00:12:04,934
But, anyway, point is that what ends up happening is that when I look at a screen, it becomes difficult for me to particularly read certain types of fonts.
144
00:12:04,934 --> 00:12:13,095
And so what we've done is we've actually had a whole bunch of assistive, components that actually completely change the interface so that this way it adapts to the user.
145
00:12:13,095 --> 00:12:17,570
And the user can change however they need to, and it doesn't matter because the whole system works together.
146
00:12:17,649 --> 00:12:21,809
So if somebody has, you know, vision problems like I do, I can easily adjust it.
147
00:12:21,809 --> 00:12:23,490
I can put it on a high contrast mode.
148
00:12:23,490 --> 00:12:26,930
I can put it on all these different modes so that this way it doesn't matter if I'm color blind.
149
00:12:26,930 --> 00:12:30,370
It doesn't matter if I have any sort of visual component that's that's challenged there.
150
00:12:30,585 --> 00:12:32,024
It's also screen reader friendly.
151
00:12:32,024 --> 00:12:36,904
So this way, we can have folks who maybe have even more visual problems than I do working in there.
152
00:12:36,904 --> 00:12:47,029
And so the benefit of this tool is that it's really, really inclusive in terms of how it's able to personalize for all these individuals who may have different needs or different interests because
153
00:12:47,029 --> 00:12:49,190
maybe you don't use the tools in the same way.
154
00:12:49,190 --> 00:12:51,589
And so the system is a bit more adaptive.
155
00:12:51,750 --> 00:12:59,909
It actually is able to kind of adapt to the user, which is really important because every user really has different concerns about what matters to them.
156
00:13:01,654 --> 00:13:08,774
So, you mentioned that this system is pulling data from a number of sensors and other sources.
157
00:13:09,254 --> 00:13:19,759
Does it do kind of old fashioned statistical process control, charts, and that sort of thing, from a purely statistical analysis
158
00:13:19,759 --> 00:13:20,639
point of view?
159
00:13:20,720 --> 00:13:26,720
And can the user, basically kinda create their own views of data?
160
00:13:27,985 --> 00:13:28,465
Yeah.
161
00:13:28,465 --> 00:13:30,705
So it on the back end, yes.
162
00:13:30,705 --> 00:13:34,144
It's it's it's it's that it's big charts and tables.
163
00:13:34,225 --> 00:13:38,065
But in the front end, the user can completely customize whatever that appearance says.
164
00:13:38,065 --> 00:13:44,809
And that goes for all of our tools, including probably one of our most essential tools, which is DAISOPs.
165
00:13:44,889 --> 00:13:47,049
It's a learning tool that happens daily.
166
00:13:47,289 --> 00:13:54,649
And and with that tool, they're able to keep up to date with whatever new thing that is going on in our in our building.
167
00:13:55,524 --> 00:13:58,004
We'll have our trainings, we'll have that element in there.
168
00:13:58,245 --> 00:14:05,205
And there, again, that adaptive approach in terms of how the actual appearance works for that user is really front and center.
169
00:14:05,205 --> 00:14:08,165
Pretty very quickly, you can click on it and adjust it as you need to.
170
00:14:09,820 --> 00:14:14,220
Any other ways that, AI is being used by your clients?
171
00:14:16,779 --> 00:14:23,419
So with AI being used to our clients, I can't explicitly say what some of our clients are doing behind the doors.
172
00:14:23,419 --> 00:14:29,554
But but with us, what happens, we have a number of different ways we utilize AI with our clients.
173
00:14:29,634 --> 00:14:39,899
One is that anyone doing a project with us, all of our since all of our, all of our operators are trained in use of AI, they can better understand
174
00:14:39,899 --> 00:14:43,339
where in the workflow AI needs to be integrated.
175
00:14:43,339 --> 00:14:46,059
And the thing is that's, think, one of the more difficult components.
176
00:14:46,059 --> 00:14:49,019
When I have discussions with a lot of folks, they really don't know where AI fits.
177
00:14:49,019 --> 00:14:53,259
Most of the time, they're like, oh, I know we have this great generative AI tool, but, like, what does it do?
178
00:14:53,259 --> 00:14:54,059
Where do I put this?
179
00:14:54,514 --> 00:15:01,154
And so what we've done is we've really given the operators a really good understanding as to how the tools can work.
180
00:15:01,154 --> 00:15:03,875
This way, when they see a project, they go, I know where to put AI.
181
00:15:03,875 --> 00:15:04,835
Let's put it here.
182
00:15:04,914 --> 00:15:05,634
Let's put it here.
183
00:15:05,634 --> 00:15:06,195
Let's put it here.
184
00:15:06,440 --> 00:15:11,959
This will help reduce the overall burden on our workflow, help make sure that we can get it done in a lot less time.
185
00:15:11,959 --> 00:15:13,799
And I think that's an important component.
186
00:15:13,799 --> 00:15:21,720
It's really a combination of our critical thinking skills from the operator as well as understanding where to put the AI tools along the workflow for our clients.
187
00:15:21,965 --> 00:15:32,125
And then also for any clients who come in and do a training on our site, we offer both, DISOPs, which is a competency assessment tool,
188
00:15:32,605 --> 00:15:42,709
to any training that happens on our site, as well as our survey tool, Compare, which essentially is a qualitative, survey that doesn't really
189
00:15:42,709 --> 00:15:52,309
concern itself so much as as the idea is, like, a normal survey where it asks similar questions, but instead actually looks to to analyze the answers because that's really what matters in a qualitative
190
00:15:52,309 --> 00:15:52,629
approach.
191
00:15:52,944 --> 00:15:59,105
And so those tools are ones we offer to any client who comes on-site, who wants to do a training, who wants to do a project.
192
00:15:59,105 --> 00:16:06,464
Because the DISOPs also allow for something called, you know you know, something that's really familiar to a lot of us in in management, which is risk management.
193
00:16:06,929 --> 00:16:16,450
Because the benefit of DISOPs is that you assess the competency of the individual as they go, you assess the competency as they learn, and you get a better understanding as to where their strengths and
194
00:16:16,450 --> 00:16:24,075
weaknesses are in a way that normal, you know, read and understand SOPs just probably doesn't capture.
195
00:16:25,675 --> 00:16:30,315
I'm not saying it's not the industry standard, but I'm just saying maybe it's not the best.
196
00:16:31,035 --> 00:16:32,554
Maybe there are better approaches.
197
00:16:32,554 --> 00:16:37,195
I mean we were ready to we learned to cheat on multiple choice tests.
198
00:16:37,670 --> 00:16:43,029
So this is non cheating because the answers, there is no good answer or bad answer.
199
00:16:43,029 --> 00:16:47,509
It statistically compares it to the expert answer.
200
00:16:47,590 --> 00:16:58,245
You actually can assess the real skills and also see how you can position, for example, an employee in a work environment
201
00:16:58,564 --> 00:17:09,640
to handle the risk and who are to be the one who are going to be able to handle the risk, who are the ones who can be promoted based on their knowledge more
202
00:17:09,640 --> 00:17:14,759
into supervisory position rather than basing it on other elements.
203
00:17:14,759 --> 00:17:25,015
And also, a bunch of us, our family have that we have perfect training records and yet there are human errors in a production.
204
00:17:25,975 --> 00:17:31,815
We can read out the sources before actually the losses occur due to this human error.
205
00:17:31,815 --> 00:17:38,169
So this is where we see a big advantage for the ISOP that it gives you a transparent tool.
206
00:17:38,329 --> 00:17:42,809
It also potentially gives you a transparent tool to go for interviews.
207
00:17:42,809 --> 00:17:48,650
People are switching jobs, and this could be a portable record that they can bring with themselves.
208
00:17:48,650 --> 00:17:58,825
So instead of lying on the resume that I did this or I did that, and I've just seen it from far away, people can actually the new employers can actually see what they did
209
00:17:58,825 --> 00:18:02,184
and how competent they are in certain processes.
210
00:18:02,184 --> 00:18:09,730
So and ultimately, this reduces the human errors, which for some of the pharma companies can cost hundreds of millions a year.
211
00:18:09,809 --> 00:18:10,130
Yes.
212
00:18:10,130 --> 00:18:20,345
And I mean, maybe not lying, but maybe some people inflate maybe what they did or something and understanding their competency directly in a traceable, transparent record
213
00:18:20,345 --> 00:18:21,704
is just a lot better.
214
00:18:22,105 --> 00:18:28,984
It is a much more portable digital certificate than, you know, you're gonna find from some other digital service, you know, system.
215
00:18:28,984 --> 00:18:30,825
We we wanted to make it useful to us.
216
00:18:31,039 --> 00:18:31,359
Mhmm.
217
00:18:31,359 --> 00:18:32,960
So we wanted to make it useful to the industry.
218
00:18:32,960 --> 00:18:36,960
We wanted to to make it so that this way, you know, people would benefit from it.
219
00:18:36,960 --> 00:18:45,119
Because, you know, we see the challenge with all of our with all of our trainees, with all of our with all of our new folks, you know, because we work a lot with colleges in terms of getting hired.
220
00:18:45,119 --> 00:18:51,325
Because as you mentioned, it can be very difficult, you know, finding the right person for the role, figuring out what they're good at, what they're not.
221
00:18:51,404 --> 00:18:56,845
These are all things that can be, you know, can be can help, this tool can help kind of delineate.
222
00:18:56,845 --> 00:19:01,440
And so and, also, DISOP stands for digitally interactive standard operating procedures.
223
00:19:01,440 --> 00:19:03,200
I probably should have mentioned that earlier.
224
00:19:03,599 --> 00:19:07,359
But it is not actually the sense of that your SOPs are digitally interactive.
225
00:19:07,359 --> 00:19:11,599
It's that this is a set of SOPs that are digitally interactive for the learning process.
226
00:19:11,599 --> 00:19:14,880
They are separate and not part of the DXP process, I should say.
227
00:19:16,644 --> 00:19:20,404
So to go off on a tangent for a second Sure.
228
00:19:20,404 --> 00:19:27,285
Was was it University of San Diego where the undergrads were a little bit deficient in math?
229
00:19:28,650 --> 00:19:39,130
And do you see this kind of an AI application, being kind of a, objective measure
230
00:19:39,130 --> 00:19:40,970
of student capability?
231
00:19:41,764 --> 00:19:50,244
And do you think it will get broader acceptance and recognition as kind of a certificate?
232
00:19:50,884 --> 00:20:01,229
I hope so because this is a transparent record and a lot of pipeline, workforce pipeline that's coming down, they can be on the spectrum.
233
00:20:01,629 --> 00:20:07,949
They don't necessarily very well, but they actually could work very well in a GMP environment.
234
00:20:08,269 --> 00:20:13,734
And this could be a transparent record for them that they're not screened out in HR.
235
00:20:13,734 --> 00:20:18,454
But I think it should replace eventually the multiple choice questions.
236
00:20:18,454 --> 00:20:26,989
So right now, hiring was based on the reputation of the institution, but doesn't necessarily mean that everybody who graduates from there are really great.
237
00:20:27,390 --> 00:20:30,750
Now there is how well you pass multiple choice questions.
238
00:20:30,750 --> 00:20:37,950
And I think by the time you get to graduate level, you already know how to cheat on those or how to beat the a little bit.
239
00:20:38,075 --> 00:20:42,234
Well, you understand well, I mean, under reading a test is a whole skill in itself.
240
00:20:42,234 --> 00:20:44,714
And, like so I I was very I was very good at it.
241
00:20:44,714 --> 00:20:50,234
And so I could I could take a test, but in a subject I've never learned and get a b on it when never having taken the class.
242
00:20:50,394 --> 00:20:52,234
Just understanding the the test itself.
243
00:20:52,234 --> 00:20:59,160
And that's not a good, you know, approach of learning it, and that's also a component of it too because it may it may have been SDSU.
244
00:20:59,160 --> 00:20:59,720
I'm not certain.
245
00:20:59,720 --> 00:21:06,680
But the the the challenge is that some subjects can be a little bit more difficult to understand that translation.
246
00:21:06,680 --> 00:21:07,000
Right?
247
00:21:07,000 --> 00:21:10,025
Because they so they got an a.
248
00:21:10,025 --> 00:21:11,144
So they got a b.
249
00:21:11,144 --> 00:21:14,424
What is an a and a b worth in this environment?
250
00:21:14,664 --> 00:21:24,750
You know, what are these grades worth in this environment when you have all of these you have both grade inflation, you have all of these components have kind of entered the picture that
251
00:21:24,750 --> 00:21:26,990
that make it very difficult to discern these things.
252
00:21:26,990 --> 00:21:28,750
And a transcript doesn't tell you that.
253
00:21:28,750 --> 00:21:30,429
A syllabus doesn't tell you that.
254
00:21:30,429 --> 00:21:37,069
A syllabus tells you what they conceptually could have learned, and it and a and a transcript tells you that they got the grade they got.
255
00:21:37,069 --> 00:21:39,984
It does not give you a really good transparent record.
256
00:21:39,984 --> 00:21:42,625
And the goal of this document is to say, hey.
257
00:21:42,785 --> 00:21:47,265
This is where this person has has strengths, weaknesses, and this is where they've shown a lot of growth.
258
00:21:47,345 --> 00:21:51,025
This is where, you know, they might be this might they might have a little more room to grow here.
259
00:21:51,509 --> 00:21:59,750
And the the goal of this is is to really make that, you know, portability, make it strong, and make it so that this way you can better understand your teams.
260
00:21:59,750 --> 00:22:01,830
It makes everyone be able to work better together.
261
00:22:01,830 --> 00:22:08,454
If we understand where someone is, it's a lot easier to communicate with them what their needs are and how they're able to work with the team.
262
00:22:08,454 --> 00:22:14,454
And specifically in pharma, we have a risk based management for our whole operation.
263
00:22:15,015 --> 00:22:25,230
So I think this could address the risk part you can put the right people who potentially in a more risky situation in your process versus putting
264
00:22:25,230 --> 00:22:33,815
somebody who shown less familiarity with the process yet into a different position.
265
00:22:33,815 --> 00:22:38,855
And I think that could actually mean millions for a pharma company, putting the right people.
266
00:22:38,855 --> 00:22:47,950
So a lot of times in real operation, you find out if somebody really knows what they're supposed to know when they made a mistake and say, oops, it's expensive.
267
00:22:47,950 --> 00:22:58,109
So this way, you could know ahead of the time if they're of doing it or likely to work well, and you can start to position your employees
268
00:22:58,109 --> 00:22:58,909
based on risk.
269
00:23:00,284 --> 00:23:09,404
And I think that's for us as a nonprofit, a big advantage to see who assess who knows what.
270
00:23:09,325 --> 00:23:11,565
I think it's translate to pharma too.
271
00:23:11,640 --> 00:23:12,599
We gotta be lean.
272
00:23:12,599 --> 00:23:13,320
We gotta be lean.
273
00:23:13,320 --> 00:23:13,559
Right?
274
00:23:13,559 --> 00:23:18,839
And, you know, if if we have the lean practices, we hope for other folks to be able to learn from from our space.
275
00:23:18,839 --> 00:23:19,559
That's what we do.
276
00:23:19,559 --> 00:23:19,799
Right?
277
00:23:19,799 --> 00:23:20,839
We we test it.
278
00:23:20,839 --> 00:23:22,359
We we we're the sandbox.
279
00:23:22,519 --> 00:23:29,575
We're a space for you to play around with, try out these new tactics, these new equipment, these new these new, implementations, whatever you wanna try out.
280
00:23:29,575 --> 00:23:32,214
We want you to try it out here, see how it works.
281
00:23:32,214 --> 00:23:39,575
That's why we try to produce this information to make it ready for folks so that folks in the industry can find a way to integrate it into their own systems and say, hey.
282
00:23:40,309 --> 00:23:41,990
This is how we wanna improve.
283
00:23:42,069 --> 00:23:43,990
We wanna reduce the cost of pharmaceuticals.
284
00:23:43,990 --> 00:23:46,869
You know, my my bill to insurance total was $6,000,000.
285
00:23:46,950 --> 00:23:48,710
You know, I'm a I'm a liver transplant patient.
286
00:23:48,710 --> 00:23:55,349
I was cured of hepatitis c, you know, almost thirty years after I was, you know, because I was born with it.
287
00:23:56,204 --> 00:23:59,724
And so, you know, to me, that reducing that cost is big.
288
00:23:59,724 --> 00:24:05,484
And all the all of our efforts, all of our things here are really in that vein of trying to reduce that cost.
289
00:24:05,484 --> 00:24:08,365
And we think industry adoption is definitely an important component.
290
00:24:08,365 --> 00:24:08,924
Yes.
291
00:24:08,899 --> 00:24:19,299
And so I think battery and workforce that has 100% competency on key technologies have improved on actually every step in the manufacturing
292
00:24:19,299 --> 00:24:20,659
of what they need to do.
293
00:24:20,659 --> 00:24:28,704
Is it like measuring liquid, measuring solid, transferring liquid, transferring solid, working in our septic environment.
294
00:24:28,704 --> 00:24:35,505
So we basically broke down the basic steps, make sure they have the competency, and we assess it with the DISOP.
295
00:24:36,210 --> 00:24:46,450
And I think, provide reduction of cost for pharma, having less human error, which related losses, which is obviously
296
00:24:46,450 --> 00:24:53,494
built into the cost of the drug actually could be a big savings for everybody all around.
297
00:24:53,494 --> 00:24:57,894
And these are expenses from these losses that shouldn't occur.
298
00:24:58,695 --> 00:25:01,894
For some of the large pharma, it's hundreds of millions a year.
299
00:25:04,480 --> 00:25:13,039
Does your operation create standard deviations just like pharma manufacturing?
300
00:25:13,599 --> 00:25:14,239
Mhmm.
301
00:25:14,559 --> 00:25:15,279
Yes.
302
00:25:15,599 --> 00:25:17,119
So we work with yeah.
303
00:25:17,119 --> 00:25:20,984
We basically have a GMP system.
304
00:25:21,065 --> 00:25:24,184
Actually, all students build a system.
305
00:25:24,184 --> 00:25:34,410
So build, you know, GMP quality system, and they operate the system so they understand that in and out of how this works, which rarely is a possibility
306
00:25:34,410 --> 00:25:41,850
for that because in pharma, you get and see a fraction of the whole GMP quality system.
307
00:25:41,850 --> 00:25:52,224
And with us, it's we want to see we want to make sure that our graduates actually see the whole quality system and understand how the different
308
00:25:52,224 --> 00:25:53,664
wheels work together.
309
00:25:53,825 --> 00:25:55,825
So yes, so that's part of it.
310
00:25:55,984 --> 00:26:06,059
So where I was going with this is many times a deviation gets resolved, quote, unquote, by saying, well,
311
00:26:06,059 --> 00:26:11,339
it was operator error, and we retrained the operator.
312
00:26:11,339 --> 00:26:11,980
Yes.
313
00:26:12,299 --> 00:26:12,779
Okay?
314
00:26:12,779 --> 00:26:23,055
Now so the training was inadequate to begin with, and we're gonna retrain the operator on a training system that was inadequate
315
00:26:23,055 --> 00:26:23,934
to begin with.
316
00:26:23,934 --> 00:26:24,734
Exactly.
317
00:26:24,815 --> 00:26:27,134
That's the logic that I see here.
318
00:26:27,134 --> 00:26:27,934
But Yeah.
319
00:26:27,934 --> 00:26:35,569
Would your AI be able to kind of cry foul and say, wait a minute.
320
00:26:35,890 --> 00:26:46,134
The human error was either, okay, the individual really screwed up or the training was not good enough or
321
00:26:46,134 --> 00:26:55,015
the SOP is confusing or there's some other root cause in the manufacturing system that needs to be addressed.
322
00:26:55,015 --> 00:26:55,414
Yeah.
323
00:26:55,414 --> 00:26:58,934
Do you see AI getting us to that place?
324
00:26:59,369 --> 00:27:00,410
Oh, I mean, absolutely.
325
00:27:00,410 --> 00:27:02,169
That's exactly before.
326
00:27:02,170 --> 00:27:11,690
So so one of the things one of the first projects, you know, that because on the side, my brother and I, you know, we we we do consulting for generative AI.
327
00:27:12,025 --> 00:27:16,744
And one of the things we did was one of the first projects we did was on deviations, was around deviations.
328
00:27:17,304 --> 00:27:20,825
And the thing is, you know, this is something that we see commonly now.
329
00:27:21,065 --> 00:27:29,579
All of the big players all have AI built into looking at their deviations because it is just it's it's a great way to summarize a lot of elements and reduce a lot of the the the hours.
330
00:27:29,579 --> 00:27:32,139
It's a pretty low hanging fruit, generally.
331
00:27:32,299 --> 00:27:37,579
But the thing is that part that you touched on was is not something I hear a lot about.
332
00:27:38,044 --> 00:27:39,085
They always hear about, hey.
333
00:27:39,085 --> 00:27:41,884
I'm moving it, but they're like, ah, but we don't wanna solve the problem, though.
334
00:27:41,884 --> 00:27:45,964
That that part is a bit absent, which is unfortunate because I think it's such a missed opportunity.
335
00:27:45,964 --> 00:27:47,244
And I think you're right.
336
00:27:47,244 --> 00:27:56,210
I and I think this is one of the areas where, you know, we're going to see our when our deviations come out come into play, and we see that element that says, hey.
337
00:27:56,210 --> 00:27:57,489
This is a human error.
338
00:27:57,730 --> 00:28:02,049
We happen to have the training record that's very transparent, that's very that's very traceable test.
339
00:28:02,049 --> 00:28:02,529
Oh, look.
340
00:28:02,529 --> 00:28:07,490
Looks like they're we see right here that they're missing this specific pipetting skill.
341
00:28:07,545 --> 00:28:12,345
What we see now why this did not work, well, because they're not doing very well in this.
342
00:28:12,345 --> 00:28:18,265
Let's go back and retrain them on this specific area so we don't have to waste time with all the rest of this because this is the real weak point.
343
00:28:18,585 --> 00:28:21,819
Focus on that, and then we'll probably see a lot better results.
344
00:28:21,819 --> 00:28:26,380
Now I say probably because we haven't instituted that yet, but we will see that happen in the future.
345
00:28:26,380 --> 00:28:27,419
I'm certain of it.
346
00:28:27,500 --> 00:28:28,059
Yeah.
347
00:28:28,299 --> 00:28:28,619
Yeah.
348
00:28:28,619 --> 00:28:35,335
I mean, using AI to write a deviation that, you know, okay.
349
00:28:35,335 --> 00:28:37,974
That's kind of an efficiency tool.
350
00:28:37,974 --> 00:28:38,454
Yeah.
351
00:28:38,534 --> 00:28:41,094
But AI is statistics.
352
00:28:41,095 --> 00:28:49,095
And using AI to statistically say, you really think it's it's retraining is gonna work?
353
00:28:49,095 --> 00:28:56,169
Well, my statistics tell me that, you have a ten percent chance of being successful with that approach.
354
00:28:56,490 --> 00:28:57,769
Do you think we'll get there?
355
00:28:57,930 --> 00:28:58,490
Yes.
356
00:28:58,490 --> 00:28:59,450
Oh, absolutely.
357
00:28:59,450 --> 00:29:02,650
That's that's kind of the goal to get there.
358
00:29:02,650 --> 00:29:08,265
But the other side of it, we also train our students focusing of the language.
359
00:29:08,265 --> 00:29:10,585
So they actually have to write their SOPs.
360
00:29:11,065 --> 00:29:18,424
And we're looking at how things in an SOP could lead to human error and how we can eliminate that.
361
00:29:18,820 --> 00:29:28,980
So we're training them also to how to write SOPs and without things like, for example, if you have a sentence that has two things in it, typically the second
362
00:29:28,980 --> 00:29:39,265
part of the sentence where the error is going to happen because if you the SOP say, do this and then do that, the second part sometimes get where
363
00:29:39,265 --> 00:29:40,224
the error happens.
364
00:29:40,224 --> 00:29:47,825
So we basically write, train them how to write SOPs to avoid human errors and set it up properly.
365
00:29:48,470 --> 00:29:52,789
And then, so we're looking at that as a potential resource as well.
366
00:29:52,789 --> 00:30:02,914
And I think, yes, it's AI going to come in because AI helps us with that to sort out some of these basic elements that based on human behavior already
367
00:30:02,914 --> 00:30:11,394
shown in other industries that leads to human errors, like a nuclear power or space industries.
368
00:30:11,394 --> 00:30:11,875
Yes.
369
00:30:11,875 --> 00:30:19,519
And at the end of the day, AI can be integrated with SOPs very SOPs are kind of the language of a lot of generative AI kind of anyway.
370
00:30:19,599 --> 00:30:25,279
If you really think about the if you break down how an LM works, I mean, you're talking about the tokens, each of the words.
371
00:30:25,279 --> 00:30:25,519
Right?
372
00:30:25,519 --> 00:30:28,720
And words are things that modify words or what tokens are.
373
00:30:28,960 --> 00:30:38,875
And and so what that means in this instance is with an SOP, it's just a series of tokens that all interrelate, and you have the right vectors that associate with them, vectors being that they tie the things
374
00:30:38,875 --> 00:30:41,034
closest to one another and associated objects.
375
00:30:41,034 --> 00:30:46,990
The the idea here is that the SOP is actually a really good framework generally for an LLM to understand.
376
00:30:47,150 --> 00:30:52,589
An LLM being a large language model like like ChatGPT or any of those other tools, DeepSea.
377
00:30:52,589 --> 00:30:56,109
Any any of the tools, even if it's off or closed or open, doesn't matter.
378
00:30:56,190 --> 00:31:02,234
The idea here is that, you know, when we use a generative AI tool, which is an LLM, sorry for using all these language here.
379
00:31:02,234 --> 00:31:02,954
Don't worry.
380
00:31:02,954 --> 00:31:07,994
There'll be you can learn more with AI literacy stuff that we do with the AI COP.
381
00:31:08,474 --> 00:31:11,594
Just a plug for the ISP, AI COP for a second.
382
00:31:12,234 --> 00:31:13,514
You know, please do join us.
383
00:31:13,514 --> 00:31:14,794
We are working more with AI literacy.
384
00:31:14,794 --> 00:31:15,194
It's great.
385
00:31:16,490 --> 00:31:22,329
But the idea here being that with SOPs, they're a prime opportunity as well.
386
00:31:22,730 --> 00:31:29,529
And maybe the way they're designed, maybe whether they're structured, like you're saying that order isn't necessarily the best way, maybe there's a way to make that better.
387
00:31:29,845 --> 00:31:37,525
And one of the nice things about this is not only with, with these tools can you suddenly make sweeping changes to different SOPs.
388
00:31:37,605 --> 00:31:40,325
You can so, obviously, you're taking them out of your quality system.
389
00:31:40,325 --> 00:31:42,005
You're not gonna do this live in a real system.
390
00:31:42,005 --> 00:31:52,299
But, you know, in a in a in a in a more of a sandbox environment, you can really kind of, you know, tweak these SOPs to to protect them in a way that is you know, would take a long,
391
00:31:52,299 --> 00:31:54,859
long time without generative AI.
392
00:31:55,340 --> 00:32:00,779
You know, we've been experimenting with a few different ways to look at SOPs generally that are not part of that quality system.
393
00:32:00,974 --> 00:32:06,095
So this way, we can see them from a learning standpoint, and we're wondering how that will eventually translate.
394
00:32:06,335 --> 00:32:11,214
And, you know, we'll talk more about that probably near the end of the year once we've got a lot more data on that.
395
00:32:11,615 --> 00:32:14,609
But, you know, for now, this is where we are.
396
00:32:14,609 --> 00:32:24,769
Reading these issues from a pharmaceutical process and looking at the training record, looking at what could lead
397
00:32:24,769 --> 00:32:31,134
to error in an SOP and potentially DCR the SOP and upgrade it to less human errors.
398
00:32:31,134 --> 00:32:34,014
Think AI is a great tool for that.
399
00:32:34,575 --> 00:32:37,375
So let me ask two related questions.
400
00:32:37,375 --> 00:32:42,255
They both are around the subject of visual or visualization.
401
00:32:43,130 --> 00:32:50,809
One, are you looking at how to integrate video into an SOP?
402
00:32:51,210 --> 00:32:58,025
And two, are you using vision systems to assess operator competency?
403
00:32:59,865 --> 00:33:02,904
So vision is an interesting situation.
404
00:33:02,904 --> 00:33:10,664
So one thing that we've been attempting to do with sensors and camera is really understand how someone is meeting the SOP.
405
00:33:10,859 --> 00:33:18,859
And that's something that we've been kind of working on for a long time, I think it's gonna take a a little while yet to really understand exactly how somebody someone is moving or is doing something and
406
00:33:18,859 --> 00:33:24,940
how that matches an SOP because of the way people are physically positioned, and there are many other factors involved in that.
407
00:33:24,940 --> 00:33:28,565
So that's why it's gonna take a little bit longer to better understand that element.
408
00:33:28,804 --> 00:33:38,325
But there are obvious other ones, like having a more more kind of digital SOP where it is it does, like you're saying, have a have video.
409
00:33:38,325 --> 00:33:41,524
Say, oh, well, this is something we don't do very often, but it's very critical.
410
00:33:41,640 --> 00:33:51,880
Here's a video of an operator exactly going through with it and explaining step by step along with that the SOP already does, but reiterating it again and showing it as they are doing it right
411
00:33:51,880 --> 00:33:52,359
through there.
412
00:33:52,359 --> 00:34:00,894
And I think we've seen, we've worked with with groups doing, AR as well on that approach and saying how this works with AR, which is useful.
413
00:34:01,214 --> 00:34:08,255
However, the the challenge I think with, you know, with with VR and with AR is, again, the certain limitations of of the tool itself.
414
00:34:08,255 --> 00:34:08,494
Right?
415
00:34:08,494 --> 00:34:11,614
So with with VR, it's a matter of, like, I can't use VR.
416
00:34:11,614 --> 00:34:18,110
Like, I'm too my my my eyes are such that, you know, it is very difficult for me to be on VR for a time.
417
00:34:18,269 --> 00:34:21,150
And I have colleagues who get vertigo very quickly on them as well.
418
00:34:21,150 --> 00:34:23,070
So it it it does have limitations.
419
00:34:23,389 --> 00:34:28,844
That being said, watching the video or seeing a video on an iPad while you're there in there and saying, okay.
420
00:34:28,844 --> 00:34:30,125
Well, this is what I need to do.
421
00:34:30,125 --> 00:34:31,164
I can see that.
422
00:34:31,164 --> 00:34:36,605
Or having a tool you can kinda show up over the iPad or the augmented reality situation, that's really amazing.
423
00:34:36,900 --> 00:34:38,179
Being able to kinda say, okay.
424
00:34:38,179 --> 00:34:42,579
Well, here's where my next steps are following that SOP along those ways.
425
00:34:42,659 --> 00:34:43,619
Phenomenal.
426
00:34:44,099 --> 00:34:54,235
Something that we definitely will look at once we have, you know, more operations going in our in our actual in in our clean rooms because at the end of the day, until that happens, we're
427
00:34:54,235 --> 00:34:55,195
not on that.
428
00:34:55,195 --> 00:34:57,275
We we can't produce data or check on it.
429
00:34:57,275 --> 00:35:04,315
But in terms of our training lab, where we are there, there, you know, it's we're mostly looking at how are their pipetting skills?
430
00:35:04,315 --> 00:35:05,594
How are they working under the hood?
431
00:35:05,869 --> 00:35:07,630
How are they working at the microscope?
432
00:35:07,630 --> 00:35:11,949
How are they how are they, you know, working in these more more traditional laboratory settings?
433
00:35:11,949 --> 00:35:13,150
So we look at it there.
434
00:35:13,550 --> 00:35:18,829
We're looking at both the video that's, being taken, and there they can kind of set up a a sample video.
435
00:35:18,829 --> 00:35:18,989
Yeah.
436
00:35:19,985 --> 00:35:20,864
So I think Alright.
437
00:35:20,864 --> 00:35:30,945
Let's question the the videos could be distracted, but I think, you know, learning in the first place and have the training and the training completed so
438
00:35:30,945 --> 00:35:32,625
they're competent with the SOP.
439
00:35:32,625 --> 00:35:34,720
I think it's kind of the good way to go.
440
00:35:34,720 --> 00:35:35,200
Yes.
441
00:35:36,160 --> 00:35:44,320
So let's shift the focus a little bit to Open BioPharma and its commercial endeavors.
442
00:35:44,320 --> 00:35:45,599
How are you funded?
443
00:35:45,599 --> 00:35:48,079
Is it public, private or a mix?
444
00:35:49,144 --> 00:35:53,385
So it's a Open Biopharma is a 501C nonprofit.
445
00:35:53,385 --> 00:35:58,025
So partially, we are funded by participating in grants.
446
00:35:58,025 --> 00:36:01,945
So we have workforce development grants and other grants with partners.
447
00:36:01,945 --> 00:36:05,320
We partner with community colleges, universities.
448
00:36:05,559 --> 00:36:14,920
We partners also across the industry, with Texas, A and M and North Carolina, and they're applying for different workforce development grants.
449
00:36:14,920 --> 00:36:25,135
So all one of our core thing to make sure we develop the future workforce, and that workforce is going to reduce these human error elements and move
450
00:36:25,135 --> 00:36:29,934
some of the new technologies potentially faster into the pharma space.
451
00:36:29,934 --> 00:36:31,135
So that's one of our source.
452
00:36:31,430 --> 00:36:35,910
We have other grant sources too, then we work as a training facility.
453
00:36:35,910 --> 00:36:39,750
So we have other companies come and train or we co train with them.
454
00:36:39,750 --> 00:36:42,630
We develop our own training program.
455
00:36:43,190 --> 00:36:52,074
And then we also working since we had a hands on, you know, on the job training, we bring in projects from companies.
456
00:36:52,074 --> 00:36:52,635
Mhmm.
457
00:36:52,795 --> 00:36:53,994
Project based learning is a big one.
458
00:36:53,994 --> 00:36:57,675
Project based learning, and we bring in projects from from companies.
459
00:36:57,755 --> 00:36:59,755
And they sponsor the training.
460
00:37:00,170 --> 00:37:07,690
For example, we agreed that this six months, the one year project, they sponsor the training and we deliver.
461
00:37:07,690 --> 00:37:13,849
And this is all for our students is they have capstone projects a lot of times.
462
00:37:14,675 --> 00:37:17,795
And we accomplished actually pretty good.
463
00:37:17,795 --> 00:37:26,994
So there are products that's going out of this process that students has developed and helping the cell and gene therapy industry soon.
464
00:37:27,179 --> 00:37:27,739
Pretty big win.
465
00:37:27,739 --> 00:37:29,819
One market and it's a pretty big win.
466
00:37:29,819 --> 00:37:41,105
And for them to have also an experience that they're developing product, we're also helping small companies to translate processes or process development into manufacturing and
467
00:37:41,105 --> 00:37:43,264
try to fill that gap.
468
00:37:43,264 --> 00:37:52,864
And this is the gap where CDMOs are not making money, but doing it the wrong way can cause the company their life or could cause their investment.
469
00:37:52,864 --> 00:37:55,824
So we're trying to help us a nonprofit in this early space.
470
00:37:56,380 --> 00:37:58,699
Too, and we're bringing in several projects.
471
00:37:58,699 --> 00:38:00,460
And then we have industry partners.
472
00:38:00,460 --> 00:38:06,139
We're working with an application data, sales force training and other application.
473
00:38:06,139 --> 00:38:16,525
And then we started also series of conferences or mini symposia around CMC.
474
00:38:16,525 --> 00:38:18,204
We found a great need for that.
475
00:38:18,204 --> 00:38:23,484
We had the first CMC conference in December, and we had close to 100 participants.
476
00:38:24,039 --> 00:38:29,880
And so the facility was packed actually, and we continue next one in April.
477
00:38:29,880 --> 00:38:36,440
So we will continue the CMC because we see the need that this is where there's a huge gap.
478
00:38:36,994 --> 00:38:37,554
Yeah.
479
00:38:37,554 --> 00:38:39,315
And we also host.
480
00:38:39,554 --> 00:38:48,594
So we so if anyone has any training that they wanna, you know, derisk and and move off their site, you know, obviously, we have the space for it both in the training lab and then the and then the CNC and
481
00:38:48,594 --> 00:38:52,289
clean room spaces as well as our training, you know, room itself.
482
00:38:52,289 --> 00:38:54,210
Like I said, upstairs, dry lab, wet lab.
483
00:38:54,210 --> 00:38:55,890
We have a lot of different options up there.
484
00:38:56,130 --> 00:38:58,210
So we host trainings from our from organizations.
485
00:38:58,210 --> 00:38:59,969
That's that's another source of revenue.
486
00:38:59,969 --> 00:39:01,730
We also put on our own trainings.
487
00:39:01,730 --> 00:39:09,324
We actually have a have an AI workshop series that we've launched, and we're gonna be having some different AI workshops throughout the year.
488
00:39:09,405 --> 00:39:11,324
We'll be having a number of number of ones.
489
00:39:11,324 --> 00:39:12,925
We have a few codeveloped ones.
490
00:39:12,925 --> 00:39:16,844
We've designed them for specific agencies and then kind of expanded them.
491
00:39:16,844 --> 00:39:18,045
We have one that we designed for CIRM.
492
00:39:18,860 --> 00:39:24,140
That's specifically a cell workshop therapy that we give off that we that we we do every year.
493
00:39:24,300 --> 00:39:33,664
And we'll be actually preparing it for industry as well because we've in conversation, it seems like a lot of folks would really like to better understand what cell and gene therapy, like, is if you're
494
00:39:33,664 --> 00:39:40,304
if you're in the sales or if you're in vision development or in that space where you just kinda entered into it and you're like, well, I'm from other areas of biopharma or life sciences, but I don't know
495
00:39:40,304 --> 00:39:41,664
what cell and gene therapy is.
496
00:39:41,664 --> 00:39:51,710
And this way, can get a they can get an immersive two week workshop where they get to do all the fun stuff with with some different I forget what cells they work with because I I
497
00:39:51,710 --> 00:39:52,349
don't work with it too many.
498
00:39:52,590 --> 00:39:54,349
Work with IPSCs.
499
00:39:54,349 --> 00:39:56,670
We work with Amazon cameras, stem cells.
500
00:39:56,670 --> 00:39:58,349
We work with CAR T cells.
501
00:39:58,349 --> 00:39:59,390
So we have T cells.
502
00:39:59,390 --> 00:40:00,590
We have microphages.
503
00:40:00,590 --> 00:40:01,954
We have all kinds of cells.
504
00:40:01,954 --> 00:40:02,114
Yeah.
505
00:40:02,114 --> 00:40:04,195
So it's a lot of fun to work around and better understand.
506
00:40:04,195 --> 00:40:04,515
Yeah.
507
00:40:04,515 --> 00:40:07,394
Those are less fun, I have to say, generally speaking.
508
00:40:07,394 --> 00:40:09,954
More fun in a laboratory setting when it's not around you.
509
00:40:10,594 --> 00:40:14,594
But, yeah, so it we have a very diverse set of different ways of funding.
510
00:40:14,594 --> 00:40:16,355
We also obviously accept donations.
511
00:40:16,440 --> 00:40:17,000
You know?
512
00:40:17,000 --> 00:40:25,880
So if you ever see our if you ever see on our on our on our on our Zephy page where we're we have a training on there, there is a little donation button you can add at the bottom as well.
513
00:40:26,679 --> 00:40:32,344
So as a nonprofit, what inspires you to do this work?
514
00:40:33,784 --> 00:40:36,664
I mean, I came from medicine.
515
00:40:36,744 --> 00:40:40,985
So I'm originally trained as an MD as a pediatric surgeon.
516
00:40:41,224 --> 00:40:44,344
And I had kids with cancer, and my box was empty.
517
00:40:44,559 --> 00:40:48,719
So I started out first on the research side, got my PhD.
518
00:40:48,960 --> 00:40:52,960
And then I realized that research is a bunch of really great ideas.
519
00:40:52,960 --> 00:40:56,239
But until it's not manufactured, it's nothing for a physician.
520
00:40:56,985 --> 00:41:07,465
So I actually, throughout my career, I worked in different aspects of this, worked on building vaccine facility, registration of
521
00:41:07,704 --> 00:41:16,719
medical device, overseeing as quality clinical studies and appreciated where the gaps are.
522
00:41:16,719 --> 00:41:23,280
And as a nonprofit, we would like to be a community resource to solve some of these gaps and difficulties for the industry.
523
00:41:23,765 --> 00:41:26,324
And we have the luxury of doing that.
524
00:41:26,324 --> 00:41:33,925
So when companies are producing and they're under the gun to get things out, they, a lot of times, don't have the time to do this.
525
00:41:34,405 --> 00:41:43,829
And we think that by providing the trained employees who already have the right thinking and the right skills, it can move into the industry.
526
00:41:43,829 --> 00:41:54,085
And we feel that filling some of these gaps on the workforce side and also providing the opportunity for small companies to de risk their
527
00:41:54,085 --> 00:41:59,525
CMC operation and be able to obtain financing.
528
00:41:59,845 --> 00:42:06,805
These are great community resources and we're also going to move new technologies into medicine, have medicines for patients.
529
00:42:06,929 --> 00:42:10,210
So these are the fun part that excites me.
530
00:42:10,210 --> 00:42:20,050
And then see the young generation make a change because I think we our generation already have a routine of how we do things.
531
00:42:20,050 --> 00:42:24,175
So I think to move the needle is gonna be the job for the young generation.
532
00:42:24,494 --> 00:42:27,215
And we want to make sure that we train them well.
533
00:42:27,295 --> 00:42:27,614
Yes.
534
00:42:27,614 --> 00:42:37,750
And when when Susan came to you with this idea of what what Open Biopharma would be, you know, I at the time I had you know, ten years ago, I was very
535
00:42:38,710 --> 00:42:39,429
dying.
536
00:42:40,070 --> 00:42:45,269
I was an end stage liver disease at that point for almost six years, maybe seven years about that point.
537
00:42:45,589 --> 00:42:49,269
And, you know, I I was dying from hepatitis c, and I contracted at birth.
538
00:42:49,269 --> 00:42:51,034
And I've been on five different treatments.
539
00:42:51,034 --> 00:42:53,835
And I had been that that small percentage that it didn't work for.
540
00:42:54,234 --> 00:42:56,714
And, you know, my bill to insurance total was massive.
541
00:42:56,714 --> 00:43:03,674
You know, by the time I got my liver transplant and was eventually cured in 2017, you know, a bill to insurance total was $6,000,000.
542
00:43:04,059 --> 00:43:04,859
That's difficult.
543
00:43:04,859 --> 00:43:06,219
Now my family didn't pay that.
544
00:43:06,219 --> 00:43:14,940
My family paid about 350,000, maybe about $450,000 out of pocket, but that's still a hefty amount of money over a life period.
545
00:43:15,179 --> 00:43:17,659
And that's really difficult, you know, to bear.
546
00:43:17,659 --> 00:43:24,885
It's difficult to to see, you know, how those things even my dog agrees with that, you know, if you could hear him barking in the background right now.
547
00:43:24,885 --> 00:43:31,605
So the thing is that this was this was a this was a challenge, you know, for me personally.
548
00:43:31,684 --> 00:43:40,239
And, you know, the fact that my life had been saved, you know, Because when I was diagnosed at 13 in '99, there was no treatment for there was no cure for hepatitis c.
549
00:43:40,239 --> 00:43:41,680
Was a treatment, but it sucked.
550
00:43:42,000 --> 00:43:43,119
It was it was a year long.
551
00:43:43,119 --> 00:43:44,000
It was very rough.
552
00:43:44,000 --> 00:43:54,224
And, you know, to have a curative treatment happen in your lifetime, to be cured of a terminal condition in your lifetime, to be saved by so many different processes,
553
00:43:54,305 --> 00:43:57,425
and then to see all of the components within that were involved.
554
00:43:57,425 --> 00:43:58,865
And it was so complex.
555
00:43:58,945 --> 00:44:09,009
And, really, just I'm here by by luck, by chance, by, by by will, by by by the by by such a combination of forces that I think that that shouldn't have to
556
00:44:09,009 --> 00:44:13,250
happen, you know, for everyone to be saved from such a circumstance.
557
00:44:13,250 --> 00:44:23,835
And so I think that one of the things that really drives us here is making sure that more people have access to life saving and life and quality of life improving medications.
558
00:44:24,235 --> 00:44:26,795
Because the thing is they exist out there.
559
00:44:26,875 --> 00:44:30,635
There are amazing things that are being invented and created all the time.
560
00:44:30,969 --> 00:44:35,610
And our industry has shown that you can improve and save so many lives.
561
00:44:35,930 --> 00:44:44,969
And I think that it is just a matter of getting the right alignment of things to make those costs and make that accessibility really happen.
562
00:44:45,324 --> 00:44:47,324
And I think that's really a lot of what we try to do here.
563
00:44:47,324 --> 00:44:57,565
We try to make things so this way the industry can have the opportunity to align itself better, to be a bit more smooth operating in terms of, you know, getting from, you know, getting from school to
564
00:44:57,565 --> 00:45:07,059
working, having that efficiency gain, having the the the access to knowledge of production because we are pretty secretive in in the pharmaceutical industry and trying to share secrets.
565
00:45:07,059 --> 00:45:08,820
We don't we try not to do that very often.
566
00:45:08,900 --> 00:45:10,579
It is what we're kind of known for.
567
00:45:10,579 --> 00:45:11,300
We have a process.
568
00:45:11,300 --> 00:45:11,780
You're like, no.
569
00:45:11,780 --> 00:45:12,820
This is my process.
570
00:45:12,820 --> 00:45:16,295
And, you know, that can be difficult for advancement.
571
00:45:16,295 --> 00:45:26,295
And so we wanna be a space, a place for people to be able to sandbox and try those things, to try the new projects, to be innovative in a way that maybe, you know, helps save and improve some
572
00:45:26,295 --> 00:45:27,094
lives in the future.
573
00:45:27,349 --> 00:45:31,910
And we just wanna be a part of this amazing industry that helps save and improve lives.
574
00:45:31,989 --> 00:45:34,949
I would like to add something really quickly.
575
00:45:34,949 --> 00:45:38,630
So I, you know, I got the training as a as a surgeon.
576
00:45:39,349 --> 00:45:49,025
And this is an industry where a lot of time employees coming out of universities and they're trained by people who never worked in the industry.
577
00:45:49,344 --> 00:45:51,105
They don't understand the problem.
578
00:45:51,105 --> 00:45:56,710
So it's very hard for them to actually train on how to solve some of the problems.
579
00:45:57,030 --> 00:46:07,030
So for me, the training aspects that I received during residency, that instead of trained in a classroom by people who never did surgery,
580
00:46:07,715 --> 00:46:17,235
trained in a surgical suite by surgeons, why I am assisting and helping with surgery is actually, I think, a very critical thing.
581
00:46:17,235 --> 00:46:27,430
And here in this industry, we're making medications and things like insulin gene therapy that goes from $05,000,000 to 5,000,000 a dose.
582
00:46:28,390 --> 00:46:34,710
And yet people are trained in classroom by faculty who potentially never did this.
583
00:46:35,030 --> 00:46:45,134
So I think changing the training and put them into a situation where they can learn by working with experts and all
584
00:46:45,134 --> 00:46:53,179
of our faculty are coming from industry, I think, make a difference of what kind of workforce we develop for the future.
585
00:46:53,179 --> 00:47:03,355
So I would like to see that students who graduate from university, even with PhDs, come and work in this kind of environment for
586
00:47:03,355 --> 00:47:13,675
six months a year as part of their rotation because I think the translation and the CMC part of this new discovery is going to actually change
587
00:47:13,675 --> 00:47:19,730
if we can incorporate something into as I said, even into the scientist training.
588
00:47:19,730 --> 00:47:24,130
So they understand that how that idea can become a product.
589
00:47:24,130 --> 00:47:25,809
And I think there's a huge gap there.
590
00:47:25,809 --> 00:47:37,355
So these are the things that we would like to feel and be more like the, not the university, but the residency space for the workforce?
591
00:47:37,914 --> 00:47:39,914
Well, thank you very much.
592
00:47:40,394 --> 00:47:50,599
In summary, we've heard today about Open Biopharma, which focuses on operator and technician training
593
00:47:50,840 --> 00:47:55,559
for pharmaceutical laboratories and manufacturing operations.
594
00:47:56,440 --> 00:48:02,855
And you also do special projects, with a GMP facility.
595
00:48:03,175 --> 00:48:11,815
So you're quite the resource, located North Of San Diego, but, you've you've got clients from across the country.
596
00:48:12,295 --> 00:48:16,680
So I'd like to thank Susan and Richard for their time with us today.
597
00:48:17,239 --> 00:48:24,920
It is great to hear about nonprofit efforts to train the pharmaceutical manufacturing and support personnel of the future.
598
00:48:25,594 --> 00:48:32,315
That brings us to the end of another episode of the ISPE podcast, shaping the future of pharma.
599
00:48:32,635 --> 00:48:41,559
Please be sure to subscribe so you don't miss future conversations with the innovators, experts and change makers driving our industry forward.
600
00:48:42,119 --> 00:48:52,420
On behalf of all of us at ISPE, thank you for listening, and we'll see you next time as we continue to explore the ideas, trends and people shaping the
601
00:48:52,420 --> 00:48:53,620
future of pharma.
