1 00:00:00,000 --> 00:00:20,960 But I guess, see your name is H, but I'll introduce myself and then maybe we can go 2 00:00:20,960 --> 00:00:22,960 around and introduce each other. 3 00:00:22,960 --> 00:00:24,680 So ourselves. 4 00:00:24,680 --> 00:00:28,600 So my name is Keegan and I founded Canonlytics. 5 00:00:28,600 --> 00:00:34,560 So I've been working in the cannabis space for three or four years now. 6 00:00:34,560 --> 00:00:40,720 Started as a laboratory analyst, found clever ways to collect data and save people a lot 7 00:00:40,720 --> 00:00:41,720 of time. 8 00:00:41,720 --> 00:00:49,480 And, you know, now I'm just trying to primarily help laboratories just do cannabis testing 9 00:00:49,480 --> 00:00:53,320 better, simply, easily. 10 00:00:53,320 --> 00:00:58,160 And I found that a lot of people have an interest in data analytics. 11 00:00:58,160 --> 00:01:01,280 So my background is in economics. 12 00:01:01,280 --> 00:01:05,040 So in a comparative advantage of mine. 13 00:01:05,040 --> 00:01:09,680 So I just thought I would just share what I know and see if there's other people that 14 00:01:09,680 --> 00:01:10,840 are interested in. 15 00:01:10,840 --> 00:01:13,060 There's quite a few. 16 00:01:13,060 --> 00:01:20,600 So it's always fun to talk. 17 00:01:20,600 --> 00:01:21,600 Let's go next. 18 00:01:21,600 --> 00:01:30,200 I guess Paul, would you want to just, I guess, reintroduce yourself real quick? 19 00:01:30,200 --> 00:01:31,200 Sure. 20 00:01:31,200 --> 00:01:32,200 Yeah. 21 00:01:32,200 --> 00:01:33,200 My name is Paul. 22 00:01:33,200 --> 00:01:39,000 I'm a data scientist with a major automotive manufacturer in Michigan. 23 00:01:39,000 --> 00:01:42,760 I'm just wrapping up a master's program in data science. 24 00:01:42,760 --> 00:01:43,760 I'm getting a little bit of feedback. 25 00:01:43,760 --> 00:01:47,760 I don't know if that's coming from somebody's speaker. 26 00:01:47,760 --> 00:01:53,200 So I'm wrapping up this master's program in data science, and I was looking for a project 27 00:01:53,200 --> 00:01:59,520 sponsor and Keegan was nice enough to agree to help me with my graduate project. 28 00:01:59,520 --> 00:02:05,320 It's on some retail analytics in the cannabis industry. 29 00:02:05,320 --> 00:02:07,600 So yeah, and Charles has been helping as well. 30 00:02:07,600 --> 00:02:16,920 So yeah, that's my story. 31 00:02:16,920 --> 00:02:17,920 I hear you Keegan. 32 00:02:17,920 --> 00:02:23,120 Sorry, I was on mute for the feedback. 33 00:02:23,120 --> 00:02:32,920 So Charles is helping analytics with an interesting project right now. 34 00:02:32,920 --> 00:02:38,760 So we're sprinting on a laboratory information management system. 35 00:02:38,760 --> 00:02:46,960 The critical part of that is importing data from scientific instruments. 36 00:02:46,960 --> 00:02:53,120 There are a lot of these instruments are spitting out data files and the labs are collecting 37 00:02:53,120 --> 00:02:58,720 them in various means, often different means for different instruments. 38 00:02:58,720 --> 00:03:03,920 And we think there's just a simple, easy way just to collect the data. 39 00:03:03,920 --> 00:03:09,120 That way people at the lab, like the analysts, quality assurance manager, scientific director 40 00:03:09,120 --> 00:03:18,600 can review the data, issue their certificates and spend more time doing science, less time 41 00:03:18,600 --> 00:03:22,400 doing data entry, which is something computers excel at. 42 00:03:22,400 --> 00:03:27,200 So Charles, would you want to pick up from there? 43 00:03:27,200 --> 00:03:34,160 So yeah, my name is Charles and I have about 27 years of programming experience and I've 44 00:03:34,160 --> 00:03:39,480 worked in a lot of different areas and now I'm sort of transitioning into data science 45 00:03:39,480 --> 00:03:46,480 and machine learning. 46 00:03:46,480 --> 00:03:54,040 So Charles is helping with the parsing and importing of the instrument data because he's 47 00:03:54,040 --> 00:04:00,240 a good data wrangler. 48 00:04:00,240 --> 00:04:08,600 And then H, would you like to introduce yourself? 49 00:04:08,600 --> 00:04:11,240 You can pass if you'd like, we can jump in. 50 00:04:11,240 --> 00:04:12,240 I'll go ahead. 51 00:04:12,240 --> 00:04:13,240 Okay. 52 00:04:13,240 --> 00:04:14,680 Can you hear me all right? 53 00:04:14,680 --> 00:04:15,680 Yeah. 54 00:04:15,680 --> 00:04:16,680 Cool. 55 00:04:16,680 --> 00:04:17,680 I'm Heather. 56 00:04:17,680 --> 00:04:22,600 I'm not going to say I'm an ex-scientist, but I haven't been in the lab since like 2014. 57 00:04:22,600 --> 00:04:28,280 But since then, I have begun to use cannabis. 58 00:04:28,280 --> 00:04:31,560 So it's like the two worlds have now combined where I've been out of the lab, but I still 59 00:04:31,560 --> 00:04:38,280 have an innate interest in data science and in the cannabis industry. 60 00:04:38,280 --> 00:04:40,480 So that is why I'm here. 61 00:04:40,480 --> 00:04:43,600 Well, that's awesome. 62 00:04:43,600 --> 00:04:51,160 You could, if you want to check out the CanLytics GitHub, it's open source. 63 00:04:51,160 --> 00:04:56,880 And so I'll put the link in the chat. 64 00:04:56,880 --> 00:05:08,720 And your unique background in the lab space would probably be helpful because you'd be 65 00:05:08,720 --> 00:05:18,520 surprised there's the amount of people with know-how of how laboratories function and 66 00:05:18,520 --> 00:05:21,880 work and the workflow. 67 00:05:21,880 --> 00:05:22,880 It's limited. 68 00:05:22,880 --> 00:05:23,880 There's a shortage. 69 00:05:23,880 --> 00:05:30,760 So there's a high demand for people that understand that workflow and can articulate it. 70 00:05:30,760 --> 00:05:33,960 So yeah, we're always happy to have you, the board. 71 00:05:33,960 --> 00:05:34,960 Yeah. 72 00:05:34,960 --> 00:05:35,960 The interest is there. 73 00:05:35,960 --> 00:05:40,240 I just think that there's not too much available where I live, unfortunately. 74 00:05:40,240 --> 00:05:45,320 So yeah, that's the conflict. 75 00:05:45,320 --> 00:05:54,160 So I'm sort of visiting labs, but there's a lot of remote work that can essentially 76 00:05:54,160 --> 00:05:55,160 be done. 77 00:05:55,160 --> 00:06:06,680 So really anyone around the world who can own the repository and as a text editor and 78 00:06:06,680 --> 00:06:07,680 make contributions. 79 00:06:07,680 --> 00:06:08,680 That's a good point. 80 00:06:08,680 --> 00:06:09,680 Good point. 81 00:06:09,680 --> 00:06:13,840 Keegan, I have a quick question for you. 82 00:06:13,840 --> 00:06:19,560 In the lab testing world, can any of that lab test work be done somehow remotely? 83 00:06:19,560 --> 00:06:22,160 I was just curious if that, I don't know anything about it. 84 00:06:22,160 --> 00:06:23,800 So I was just curious. 85 00:06:23,800 --> 00:06:25,760 In an ideal world. 86 00:06:25,760 --> 00:06:30,460 So the lab space is real interesting. 87 00:06:30,460 --> 00:06:33,520 So there's amazing opportunities, right? 88 00:06:33,520 --> 00:06:35,980 So the instruments are powerful. 89 00:06:35,980 --> 00:06:39,000 So they have a lot of capabilities. 90 00:06:39,000 --> 00:06:50,360 So if you had everything set up perfectly, you could operate instruments with minimal 91 00:06:50,360 --> 00:06:52,840 effort by the analysts. 92 00:06:52,840 --> 00:07:00,680 You could get the data and you could almost view the data remotely. 93 00:07:00,680 --> 00:07:03,740 Potentially issue certificates remotely. 94 00:07:03,740 --> 00:07:11,920 You could view things like chromatographs and even potentially operate the instruments 95 00:07:11,920 --> 00:07:17,440 themselves remotely in an ideal world. 96 00:07:17,440 --> 00:07:27,360 In reality, getting the network set up at the laboratories is a huge challenge. 97 00:07:27,360 --> 00:07:34,520 So Heather can attest to this, it's hard to have everything communicating. 98 00:07:34,520 --> 00:07:45,680 So typically the microbiology department has their data stream and things coming off of 99 00:07:45,680 --> 00:07:53,760 the gas chromatographies coming in the HPLC, that data streaming somewhere. 100 00:07:53,760 --> 00:08:02,760 And then they may have like a LCMSMS doing pesticides. 101 00:08:02,760 --> 00:08:09,440 And then that may have an entirely different data stream. 102 00:08:09,440 --> 00:08:14,080 And it's for a variety of factors, right? 103 00:08:14,080 --> 00:08:17,560 You may set up your instruments in staggered times. 104 00:08:17,560 --> 00:08:20,440 You may set one up in January. 105 00:08:20,440 --> 00:08:25,840 You may set one up in April and another one up in August. 106 00:08:25,840 --> 00:08:28,560 They could be different manufacturers. 107 00:08:28,560 --> 00:08:34,240 So you may go with Agilent for your HPLC. 108 00:08:34,240 --> 00:08:45,480 And then you may have another vendor for your, if you're doing heavy metals, like an ICPMS. 109 00:08:45,480 --> 00:08:54,920 And then if you've got two different vendors, they don't have the incentive to help you 110 00:08:54,920 --> 00:08:57,200 aggregate your data. 111 00:08:57,200 --> 00:08:58,400 Yeah. 112 00:08:58,400 --> 00:09:06,120 It sounds like there's a lot of potential for, I guess, like, well, what you're doing, 113 00:09:06,120 --> 00:09:07,200 right? 114 00:09:07,200 --> 00:09:14,420 Like lab automation, bringing together everything in one source and integration, and then probably 115 00:09:14,420 --> 00:09:17,480 scaling as well, if you get past those hurdles. 116 00:09:17,480 --> 00:09:18,480 It's interesting. 117 00:09:18,480 --> 00:09:19,480 Okay. 118 00:09:19,480 --> 00:09:21,480 Well, it's, well, exactly. 119 00:09:21,480 --> 00:09:26,640 And so that's actually what Charles and I are working on really this week. 120 00:09:26,640 --> 00:09:34,880 So basically by next week, hopefully we'll have like a minimal use case in potentially 121 00:09:34,880 --> 00:09:35,880 in production. 122 00:09:35,880 --> 00:09:36,880 Oh, that's awesome. 123 00:09:36,880 --> 00:09:37,880 Congrats. 124 00:09:37,880 --> 00:09:49,640 So, well, we've got some coding ahead of us, but the idea is the data files are a little 125 00:09:49,640 --> 00:09:59,160 different instrument to instrument and maybe how the data files are output. 126 00:09:59,160 --> 00:10:06,400 But you can maybe kind of abstract that a little bit and just kind of take that into 127 00:10:06,400 --> 00:10:10,280 consideration and because ultimately you're just getting the data. 128 00:10:10,280 --> 00:10:17,480 So really you just want to get the data from all the instruments and just, okay, this instrument's 129 00:10:17,480 --> 00:10:18,480 an Agilent. 130 00:10:18,480 --> 00:10:22,680 So we'll just take that into consideration when we're parsing the data. 131 00:10:22,680 --> 00:10:27,720 Or this one's a Schumas do HPLC. 132 00:10:27,720 --> 00:10:32,200 So we need to parse the data slightly differently. 133 00:10:32,200 --> 00:10:40,400 But then once the data is parsed, let's just upload it to the database, all the data relatively 134 00:10:40,400 --> 00:10:41,400 the same. 135 00:10:41,400 --> 00:10:48,280 So we can kind of structure the data the same and get it in real time. 136 00:10:48,280 --> 00:11:02,000 So from just deep thought about the laboratory space, I think this is a particular bottleneck 137 00:11:02,000 --> 00:11:13,360 that you can shave off a huge amount of time on your turnaround time because it's essentially. 138 00:11:13,360 --> 00:11:20,800 So your instrument, you've loaded up maybe 50 to 100 samples and each sample takes 10 139 00:11:20,800 --> 00:11:23,200 minutes to run. 140 00:11:23,200 --> 00:11:34,720 So you could wait for 500 to 1000 minutes and then parse all the data and then issue 141 00:11:34,720 --> 00:11:37,680 your certificates. 142 00:11:37,680 --> 00:11:45,640 Or if you somehow have this nice and automated, sample one gets tested, maybe you get 10 samples 143 00:11:45,640 --> 00:11:52,400 tested, you can issue 10 certificates and then everything else is still running. 144 00:11:52,400 --> 00:12:01,400 So it helps smooth out your production so it's not so lumpy. 145 00:12:01,400 --> 00:12:11,240 And you can prioritize things, you can get some. 146 00:12:11,240 --> 00:12:12,240 It's amazing. 147 00:12:12,240 --> 00:12:19,360 So if you review sample one's results before the instrument's finished running, you can 148 00:12:19,360 --> 00:12:30,640 potentially run and prepare a new quality assurance check. 149 00:12:30,640 --> 00:12:31,640 This will happen. 150 00:12:31,640 --> 00:12:38,320 So if you're running edibles and you just get a wild result, you typically do a quality 151 00:12:38,320 --> 00:12:41,600 assurance check, just a sanity check. 152 00:12:41,600 --> 00:12:44,240 Okay, this is what we think it was. 153 00:12:44,240 --> 00:12:50,040 There was an analyst in it. 154 00:12:50,040 --> 00:12:53,160 Like missed pipette or some sort of. 155 00:12:53,160 --> 00:12:54,160 Right. 156 00:12:54,160 --> 00:13:04,240 So it sounds like with the, and maybe Heather knows a little bit about this as well, but 157 00:13:04,240 --> 00:13:14,080 so the opportunity for you to streamline and optimize is there, but you can't scale people 158 00:13:14,080 --> 00:13:15,080 very well. 159 00:13:15,080 --> 00:13:23,440 So from a technical expert kind of perspective, we talked about this a little last time, but 160 00:13:23,440 --> 00:13:27,200 where do you think the bottlenecks are as far as like having the right people in the 161 00:13:27,200 --> 00:13:31,480 lab and that being a bottleneck? 162 00:13:31,480 --> 00:13:37,760 Honestly, it's really just based on my experience, the finances in the lab. 163 00:13:37,760 --> 00:13:47,520 So all of this mention HPLC, LCMS, like everything I did was manual up until the very last time 164 00:13:47,520 --> 00:13:50,520 that I left. 165 00:13:50,520 --> 00:13:55,800 I hope to go back one day, but all the column work, everything, having to walk in and out 166 00:13:55,800 --> 00:14:03,080 of the, any dark room, cold room, everything is running not with an automated system that 167 00:14:03,080 --> 00:14:06,200 will even track and tell you like, oh, this is running low. 168 00:14:06,200 --> 00:14:08,560 You should add this or wash with this solution. 169 00:14:08,560 --> 00:14:10,840 There's nothing to everything was manual. 170 00:14:10,840 --> 00:14:18,880 So my, and would it, I mean, ultimately I, you know, my decision to leave academia was 171 00:14:18,880 --> 00:14:25,080 also because you have to decide, do you want to, you know, to fight for grants for the 172 00:14:25,080 --> 00:14:26,720 rest of your life? 173 00:14:26,720 --> 00:14:30,040 And the thing is that's okay if you want to do that, but I know that I didn't have what 174 00:14:30,040 --> 00:14:31,040 it took. 175 00:14:31,040 --> 00:14:38,440 And I know the impact within the lab on of the finances on the anybody working in there, 176 00:14:38,440 --> 00:14:42,760 they become competitive among each other within the lab. 177 00:14:42,760 --> 00:14:50,840 And that can affect the, I guess, the scientific growth of everybody working there. 178 00:14:50,840 --> 00:14:56,320 It's so, I don't mean to like stray off the main topic, but that's just my opinion is 179 00:14:56,320 --> 00:15:03,440 it's just literally how much money does your lab have and can they keep the people in there 180 00:15:03,440 --> 00:15:07,720 long enough so that they make something and create something, find something that's a 181 00:15:07,720 --> 00:15:08,720 breakthrough something. 182 00:15:08,720 --> 00:15:24,400 Heather hit on an interesting observation that goes to, yeah, like, well, yeah, how 183 00:15:24,400 --> 00:15:25,920 we really envision labs. 184 00:15:25,920 --> 00:15:36,080 So basically it's like, Oh, Keegan, we lost you. 185 00:15:36,080 --> 00:15:37,360 No, I'm just thinking. 186 00:15:37,360 --> 00:15:41,280 So basically you have like your method development managers. 187 00:15:41,280 --> 00:15:44,920 And so you want them like doing science and research. 188 00:15:44,920 --> 00:15:52,880 So you want them trying to optimize your methods, figure out how things are running. 189 00:15:52,880 --> 00:15:58,320 But what they end up doing is like Heather says, they have to record temperatures. 190 00:15:58,320 --> 00:16:02,440 They have to, you know, they have to make reagents. 191 00:16:02,440 --> 00:16:08,160 They have to record IDs in their lab notebooks. 192 00:16:08,160 --> 00:16:14,360 And so you end up spending, if things aren't like super efficient, then you end up spending 193 00:16:14,360 --> 00:16:23,840 a lot of time just taking measurements and doing like trivial tasks. 194 00:16:23,840 --> 00:16:31,200 And then you never have time to do any research or any science or you have limited time. 195 00:16:31,200 --> 00:16:33,800 It really cuts into your time. 196 00:16:33,800 --> 00:16:37,000 So I didn't know that. 197 00:16:37,000 --> 00:16:43,440 So I was just naively thinking that when these labs are running, they're just trying to crank 198 00:16:43,440 --> 00:16:47,660 through generating lab results and reports so they can get to the customer. 199 00:16:47,660 --> 00:16:52,440 So it's more there's actually research going on in these testing facilities. 200 00:16:52,440 --> 00:16:55,680 Well, essentially, yes. 201 00:16:55,680 --> 00:17:00,720 So like at the cannabis industry, it's cutting edge and it's competitive. 202 00:17:00,720 --> 00:17:05,040 So there's some like research articles. 203 00:17:05,040 --> 00:17:10,800 So there's like some like gold standard articles and like everybody tries to kind of base their 204 00:17:10,800 --> 00:17:14,480 methods off of these articles. 205 00:17:14,480 --> 00:17:18,280 But then you kind of run into complications. 206 00:17:18,280 --> 00:17:29,400 You realize, oh, in practice, this will gunk up your instrument. 207 00:17:29,400 --> 00:17:39,200 And so you have to, you know, take extra precautions or you have to there's like quality assurance 208 00:17:39,200 --> 00:17:46,520 steps that you can take that aren't well defined. 209 00:17:46,520 --> 00:17:50,760 So the standardization doesn't seem like it's because it's a new industry. 210 00:17:50,760 --> 00:17:57,040 There's not that kind of I'm not say a recipe or formula, but it seems like that's all kind 211 00:17:57,040 --> 00:18:01,360 of being kind of discovered and developed. 212 00:18:01,360 --> 00:18:02,680 Exactly. 213 00:18:02,680 --> 00:18:13,280 It's real like and it's it's kind of competitive, so people are kind of they like, you know, 214 00:18:13,280 --> 00:18:19,200 they they've spent a lot of time and money developing their methods. 215 00:18:19,200 --> 00:18:25,520 So like, when they're doing these things, they may break a piece in their instrument 216 00:18:25,520 --> 00:18:35,360 or they may spend a whole bunch of money on strange compounds or reagents. 217 00:18:35,360 --> 00:18:38,120 So they may like try this one reagent. 218 00:18:38,120 --> 00:18:39,880 And then they may. 219 00:18:39,880 --> 00:18:45,120 So basically, like when you're putting things on the HPLC, you use solvents. 220 00:18:45,120 --> 00:18:50,600 And then when you put things on a GC, you use carrier gases. 221 00:18:50,600 --> 00:18:53,200 And you can use different types of them. 222 00:18:53,200 --> 00:18:59,680 So you you may experiment and so you say, oh, this one gas. 223 00:18:59,680 --> 00:19:09,040 Like I'm not like a GC expert, but maybe there's hydrogen or nitrogen and various mixes. 224 00:19:09,040 --> 00:19:16,520 And then with the HPLC, I think there's like methanol and maybe acetonitrile. 225 00:19:16,520 --> 00:19:21,920 And so you may experiment and you'd say, oh, maybe methanol works better. 226 00:19:21,920 --> 00:19:25,280 Or maybe the acetonitrile works better. 227 00:19:25,280 --> 00:19:26,280 And. 228 00:19:26,280 --> 00:19:30,440 Sorry, Keegan. 229 00:19:30,440 --> 00:19:37,440 So if I guess states have different standards for their testing, right? 230 00:19:37,440 --> 00:19:44,480 So there's no standardized across the board like these 12 things that we have, we have 231 00:19:44,480 --> 00:19:50,320 a threshold and you can't pass the threshold on these 12 measures, whatever they are. 232 00:19:50,320 --> 00:19:51,320 There's nothing like that. 233 00:19:51,320 --> 00:19:53,560 This kind of the Wild West. 234 00:19:53,560 --> 00:19:59,360 Well, typically you have to pass. 235 00:19:59,360 --> 00:20:03,120 You have to like. 236 00:20:03,120 --> 00:20:05,800 Like match like standards, they said. 237 00:20:05,800 --> 00:20:10,760 So I'm not sure if I'm using the right terminology. 238 00:20:10,760 --> 00:20:16,760 Which the word proficient like proficiency tests. 239 00:20:16,760 --> 00:20:24,760 So they'll basically send you a substance and you have to measure the. 240 00:20:24,760 --> 00:20:29,360 Like the like the concentrations within a certain bound. 241 00:20:29,360 --> 00:20:36,560 So you have to be like within maybe 10 percent of what they expect or something like that. 242 00:20:36,560 --> 00:20:44,480 So you know, when everybody's getting their license and getting accredited, they basically 243 00:20:44,480 --> 00:20:49,120 have to prove that, OK, yes, we can measure pesticides. 244 00:20:49,120 --> 00:20:56,440 We can get the concentrations, you know, relatively close to what they're going to be because 245 00:20:56,440 --> 00:21:01,040 there's always a bound that your confidence bounds. 246 00:21:01,040 --> 00:21:10,840 And so with a lot of the method development is, is just getting more accurate, more consistent. 247 00:21:10,840 --> 00:21:15,600 Because then. 248 00:21:15,600 --> 00:21:19,200 When you're more accurate and more consistent, like like there's less things that are called 249 00:21:19,200 --> 00:21:20,200 like out of spec. 250 00:21:20,200 --> 00:21:21,200 So then. 251 00:21:21,200 --> 00:21:28,160 To say you like, you know, you run your your standard and it's not what you expect. 252 00:21:28,160 --> 00:21:35,760 So you see you're expecting it to come out at 80 percent THC. 253 00:21:35,760 --> 00:21:41,640 This is a rough example and your standard comes out at like 60 percent THC. 254 00:21:41,640 --> 00:21:45,320 You're like, OK, something went wrong. 255 00:21:45,320 --> 00:21:47,040 What went wrong? 256 00:21:47,040 --> 00:21:54,040 And then, you know, potentially we're going to have to maybe relook at all of these samples 257 00:21:54,040 --> 00:21:57,160 because something may have just gone wrong in the. 258 00:21:57,160 --> 00:21:58,800 I see. 259 00:21:58,800 --> 00:22:04,280 So there are there are different standards between states. 260 00:22:04,280 --> 00:22:08,360 Like what's what you have to test for and what's acceptable. 261 00:22:08,360 --> 00:22:10,880 Oregon is really different than Washington. 262 00:22:10,880 --> 00:22:17,040 And I don't know about California or Michigan or any other states, but I know between Oregon 263 00:22:17,040 --> 00:22:23,000 and Washington, there's a huge difference about what's what's allowable and what they 264 00:22:23,000 --> 00:22:24,680 have to test for. 265 00:22:24,680 --> 00:22:29,840 OK, so that that definitely complicates things, doesn't it? 266 00:22:29,840 --> 00:22:30,840 Yes. 267 00:22:30,840 --> 00:22:37,800 And like Charles said, with pesticides, there's a whole nother level of complication because 268 00:22:37,800 --> 00:22:41,800 the cannabinoids, they've got it pretty well dialed in. 269 00:22:41,800 --> 00:22:50,000 There's like a few like, you know, tricks of the trade where people can better analyze 270 00:22:50,000 --> 00:22:51,800 the cannabinoids and others. 271 00:22:51,800 --> 00:22:56,120 But generally they can measure them fairly well. 272 00:22:56,120 --> 00:23:01,240 And that's all about just everybody's just trying to do it as quickly as possible. 273 00:23:01,240 --> 00:23:03,880 So everyone's trying to just shorten their test time. 274 00:23:03,880 --> 00:23:04,880 Right. 275 00:23:04,880 --> 00:23:09,600 So you could do it a long one, maybe like 16 to 18 minutes. 276 00:23:09,600 --> 00:23:14,520 If you get a super good instrument, you can do it in like four to six. 277 00:23:14,520 --> 00:23:20,000 And then typically maybe somewhere in between there, you know, you're trying to push that 278 00:23:20,000 --> 00:23:21,920 down. 279 00:23:21,920 --> 00:23:26,480 The pesticides you're testing. 280 00:23:26,480 --> 00:23:36,040 Many some states, maybe just 12, but typically 70 plus in Canada, it's in the hundreds. 281 00:23:36,040 --> 00:23:40,320 So like 120 or so. 282 00:23:40,320 --> 00:23:44,520 And then when you're doing that, there's all these complications. 283 00:23:44,520 --> 00:23:50,600 So maybe this one compound is just notorious for false positives. 284 00:23:50,600 --> 00:23:55,280 And so if you run it your instrument one way, you'll get a false positive. 285 00:23:55,280 --> 00:23:58,760 And then you're like, okay, let's run it the other way. 286 00:23:58,760 --> 00:24:04,280 And then it's like, oh, no, that's not what we think it is. 287 00:24:04,280 --> 00:24:08,440 So there's and this is where I'm not an expert. 288 00:24:08,440 --> 00:24:11,200 No, I'm sorry. 289 00:24:11,200 --> 00:24:13,360 There's a lot going on. 290 00:24:13,360 --> 00:24:19,680 I definitely don't want to, you know, totally run a rough shot over the over the conversation. 291 00:24:19,680 --> 00:24:21,360 So I was just, it's really fascinating. 292 00:24:21,360 --> 00:24:22,720 I know there's a lot going on here. 293 00:24:22,720 --> 00:24:26,640 And I just didn't, I really don't know much about this area, especially, you know, in 294 00:24:26,640 --> 00:24:27,640 the testing space. 295 00:24:27,640 --> 00:24:28,640 So just curious. 296 00:24:28,640 --> 00:24:36,080 It's it's yeah, we don't have to get bogged down in this, but it's yeah, it's quite interesting. 297 00:24:36,080 --> 00:24:43,960 And it adds a lot of good data points that are often missing. 298 00:24:43,960 --> 00:24:53,120 Because now we like, you know, we know a lot about the actual products themselves. 299 00:24:53,120 --> 00:24:59,160 So we can kind of see, okay, well, what appeals to people. 300 00:24:59,160 --> 00:25:06,400 So this is where some some ideas are, but there's not much research, but just kind of 301 00:25:06,400 --> 00:25:13,720 seeing, okay, dude, you know, certain people gravitate to certain, you know, breakdowns 302 00:25:13,720 --> 00:25:18,560 of these compounds of chemicals and like the terpene. 303 00:25:18,560 --> 00:25:19,560 Gotcha. 304 00:25:19,560 --> 00:25:22,120 Well, thanks for sharing that. 305 00:25:22,120 --> 00:25:25,000 I, yeah, like I said, I didn't have much perspective on it. 306 00:25:25,000 --> 00:25:30,280 But now I kind of see where all the some of the new nuance creeps in here and where the 307 00:25:30,280 --> 00:25:32,800 opportunity space is. 308 00:25:32,800 --> 00:25:40,440 Yes, because, you know, like, like I said, you know, you know, I'm in essentially the 309 00:25:40,440 --> 00:25:42,280 business of data wrangling. 310 00:25:42,280 --> 00:25:47,360 And this is like, you know, Charles is going to find out quickly. 311 00:25:47,360 --> 00:25:55,600 This is hundreds, depending on how you want to count them thousands of data points. 312 00:25:55,600 --> 00:25:58,920 And so you can't necessarily get them all. 313 00:25:58,920 --> 00:26:07,000 But basically, we're just trying to collect as much data as we can and help help the labs 314 00:26:07,000 --> 00:26:09,200 essentially review the data, right? 315 00:26:09,200 --> 00:26:16,600 Because if they're spending all their time entering in data, then that's a lot less time 316 00:26:16,600 --> 00:26:24,560 that they get to spend actually reviewing it, making sure things like look like they're 317 00:26:24,560 --> 00:26:30,000 in spec and enduring a method development. 318 00:26:30,000 --> 00:26:37,720 Yeah, I could see where the lab would definitely be could potentially be a choke point right 319 00:26:37,720 --> 00:26:41,280 in the whole the whole value chain. 320 00:26:41,280 --> 00:26:47,200 Because if you have trouble getting the right people in the lab, have trouble with standards, 321 00:26:47,200 --> 00:26:52,080 you have trouble with calibrating your tests and all these things, it definitely seems 322 00:26:52,080 --> 00:26:54,880 like there could be a lot that could be improved. 323 00:26:54,880 --> 00:26:55,880 Yes. 324 00:26:55,880 --> 00:27:06,240 And just from from what I've observed, it seems that for whatever reason, it's basically 325 00:27:06,240 --> 00:27:15,440 like the maybe the like the wholesalers, it's all like the wholesalers are always just trying 326 00:27:15,440 --> 00:27:18,080 to find retailers to, you know, buy their products. 327 00:27:18,080 --> 00:27:22,000 And it seems like, okay, once they've established their relationship, then it's like, okay, 328 00:27:22,000 --> 00:27:25,560 now let's get our products tested and sold. 329 00:27:25,560 --> 00:27:31,400 And so then they basically they show up at the lab and they're like, hey, will you test 330 00:27:31,400 --> 00:27:34,080 this like we're ready to sell it. 331 00:27:34,080 --> 00:27:37,640 And then the labs like, oh, you know, we're backed up. 332 00:27:37,640 --> 00:27:44,400 Yeah, it's going to take like five days plus the weekend and or method development managers 333 00:27:44,400 --> 00:27:48,200 on vacation or, you know, what have you. 334 00:27:48,200 --> 00:27:56,320 And so then like, yeah, the wholesalers just calling up the lab saying like, hey, like 335 00:27:56,320 --> 00:27:59,320 we're trying to trying to sell our products. 336 00:27:59,320 --> 00:28:03,000 You know, we're we're results. 337 00:28:03,000 --> 00:28:04,000 Yeah. 338 00:28:04,000 --> 00:28:10,360 And so this is I just had it was not in her head. 339 00:28:10,360 --> 00:28:14,200 It sounds like she's probably suffered through some of this stuff. 340 00:28:14,200 --> 00:28:17,800 Have you had her? 341 00:28:17,800 --> 00:28:25,160 Well, yeah, I mean, I feel like sometimes there's like a dearth of terpene terpene levels 342 00:28:25,160 --> 00:28:27,720 in the flower in the state that I live. 343 00:28:27,720 --> 00:28:29,600 And it doesn't matter what the brand is. 344 00:28:29,600 --> 00:28:31,560 It's just a pattern that I've noticed. 345 00:28:31,560 --> 00:28:35,840 There are little local things like that. 346 00:28:35,840 --> 00:28:38,080 Like why going on? 347 00:28:38,080 --> 00:28:39,320 Maybe I'm overthinking it. 348 00:28:39,320 --> 00:28:41,920 You know, it could have been that gorilla glue. 349 00:28:41,920 --> 00:28:47,560 I don't know. 350 00:28:47,560 --> 00:28:54,960 So maybe so that could be a sign of basically people just rushing things to market. 351 00:28:54,960 --> 00:28:57,360 Yeah, you can tell. 352 00:28:57,360 --> 00:29:00,520 Like this is usually almost at two percent mercy. 353 00:29:00,520 --> 00:29:02,060 Why is it my why? 354 00:29:02,060 --> 00:29:07,560 Not that I wouldn't still buy it necessarily, but it's like point four percent or less than 355 00:29:07,560 --> 00:29:08,560 like it's not. 356 00:29:08,560 --> 00:29:13,360 I mean, it's the same, you know, genetically the same strain, but it doesn't hit as hard. 357 00:29:13,360 --> 00:29:19,840 So depending on what you need it for medically or otherwise, that's just, you know, something 358 00:29:19,840 --> 00:29:20,840 I'm mindful of. 359 00:29:20,840 --> 00:29:24,280 So yes, I'm laughing at my own pain. 360 00:29:24,280 --> 00:29:31,800 Well, it's interesting to observe because we've talked about this before. 361 00:29:31,800 --> 00:29:35,280 The great debate is, was does THC matter? 362 00:29:35,280 --> 00:29:41,920 And in your case, you're going beyond that and you're saying, you know, mercy matters, 363 00:29:41,920 --> 00:29:50,040 which is which is interesting because we. 364 00:29:50,040 --> 00:29:52,360 You'll hear a lot of different takes. 365 00:29:52,360 --> 00:29:59,760 So a lot of people put a lot of stock into the cannabinoids terpene data and they'll 366 00:29:59,760 --> 00:30:02,760 even base their breeding strategies on that. 367 00:30:02,760 --> 00:30:16,040 And then other people will just, you know, dismiss it and they just go by rules of thumb. 368 00:30:16,040 --> 00:30:22,520 But they are there's both successful players in both camps. 369 00:30:22,520 --> 00:30:32,680 So is there a big drive to quickly set up labs to meet the demand? 370 00:30:32,680 --> 00:30:36,160 Well, it doesn't necessarily have to be quickly. 371 00:30:36,160 --> 00:30:39,480 There is just a high demand to set up labs. 372 00:30:39,480 --> 00:30:40,480 So. 373 00:30:40,480 --> 00:30:43,480 And it takes some time. 374 00:30:43,480 --> 00:30:44,480 So maybe you're right. 375 00:30:44,480 --> 00:30:49,400 So maybe there is demand to do it quickly because people always wait too long to do 376 00:30:49,400 --> 00:30:50,400 things. 377 00:30:50,400 --> 00:30:58,040 So, for example, with the labs, you really want to get started like a year before, at 378 00:30:58,040 --> 00:31:04,600 least before. 379 00:31:04,600 --> 00:31:09,840 Your state were to get cannabis permitted. 380 00:31:09,840 --> 00:31:11,520 So like a year before. 381 00:31:11,520 --> 00:31:17,920 So if you like saw that, oh, I think. 382 00:31:17,920 --> 00:31:25,240 Whatever the next state is to permit cannabis. 383 00:31:25,240 --> 00:31:28,880 Louisiana maybe, but. 384 00:31:28,880 --> 00:31:34,040 Or rather a lot of labs are setting up in Florida. 385 00:31:34,040 --> 00:31:36,720 I think New York's not yet online. 386 00:31:36,720 --> 00:31:43,720 So I bet you there's a lot of labs that are setting up right now in New York. 387 00:31:43,720 --> 00:31:50,000 And it's not always too late to set up a new lab in a state that's already going because 388 00:31:50,000 --> 00:31:51,440 there's a high demand for it. 389 00:31:51,440 --> 00:31:56,280 But it does take some time because you're going to have to get your building, get all 390 00:31:56,280 --> 00:32:04,240 your instruments in there, get your team, then get like sort of internally validated 391 00:32:04,240 --> 00:32:06,560 to make sure like, yes, we can test things. 392 00:32:06,560 --> 00:32:12,800 Then you're going to have to get like certified by the state. 393 00:32:12,800 --> 00:32:19,400 And then, yeah, then start ironing things out, growing your team, hiring lab analysts. 394 00:32:19,400 --> 00:32:23,880 Yeah, and that's not even considering all the financing that must be involved to try 395 00:32:23,880 --> 00:32:25,240 and set something like this up. 396 00:32:25,240 --> 00:32:28,440 Must be a lot of money. 397 00:32:28,440 --> 00:32:30,120 Yes. 398 00:32:30,120 --> 00:32:41,760 So it's interesting to see people pull it off because there's some, so like there's 399 00:32:41,760 --> 00:32:52,800 an awesome example of somebody in Washington state, Jeff Doughty, Capital Analytics. 400 00:32:52,800 --> 00:33:01,040 So he basically started as a cultivator, just a small cultivator, and then bought an HPLC, 401 00:33:01,040 --> 00:33:03,040 which is not cheap. 402 00:33:03,040 --> 00:33:07,280 That could run you $40,000 to $80,000. 403 00:33:07,280 --> 00:33:10,280 That's not negligible. 404 00:33:10,280 --> 00:33:15,360 But then he just started testing cannabinoids in-house. 405 00:33:15,360 --> 00:33:19,280 And then people just were like, oh, like, and so I think he's maybe started doing a 406 00:33:19,280 --> 00:33:21,720 lot of research and development testing. 407 00:33:21,720 --> 00:33:26,480 So just doing essentially cannabinoids. 408 00:33:26,480 --> 00:33:28,720 Yeah, essentially cannabinoids. 409 00:33:28,720 --> 00:33:33,320 And then he built up essentially, well, he built up a lab. 410 00:33:33,320 --> 00:33:37,080 So then you would get a GC next. 411 00:33:37,080 --> 00:33:40,720 So you would get a GC or a little cheaper. 412 00:33:40,720 --> 00:33:50,400 There may be a real cheap one, maybe like 15, but maybe 15 to 30 or 40,000. 413 00:33:50,400 --> 00:33:55,000 And then you can start testing turkeys. 414 00:33:55,000 --> 00:34:02,560 And at the same time, you can set up your microbiology lab just to do simple testing 415 00:34:02,560 --> 00:34:06,320 of microbes and micro toxins. 416 00:34:06,320 --> 00:34:08,360 That's still going to be expensive. 417 00:34:08,360 --> 00:34:16,240 But if you get a small team, you may be able to run that for 10 or 15 or 1,000. 418 00:34:16,240 --> 00:34:22,120 So you can start a small lab if you've got a science. 419 00:34:22,120 --> 00:34:31,600 It often takes an entrepreneur and someone with a good science background that can be 420 00:34:31,600 --> 00:34:34,720 the same person I've observed. 421 00:34:34,720 --> 00:34:39,880 But it can be any team. 422 00:34:39,880 --> 00:34:48,560 But I've seen it done where they start small just doing the basic tests and then gradually 423 00:34:48,560 --> 00:34:50,320 build up some clients. 424 00:34:50,320 --> 00:34:56,920 And then they can buy the big instruments like the LCMSMS, which is going to be like 425 00:34:56,920 --> 00:34:57,920 300,000. 426 00:34:57,920 --> 00:35:04,400 And the ICP, which is 175. 427 00:35:04,400 --> 00:35:13,680 And then once you get instruments like that, you have to get expert. 428 00:35:13,680 --> 00:35:23,080 You really, really need method development managers and several chemists and several 429 00:35:23,080 --> 00:35:26,480 analysts just to make sure you can get everything done. 430 00:35:26,480 --> 00:35:31,960 So I would honestly recommend. 431 00:35:31,960 --> 00:35:33,840 Well I'm not sure what I would recommend. 432 00:35:33,840 --> 00:35:40,240 So you do see labs just starting up just with everything, just going full speed. 433 00:35:40,240 --> 00:35:48,240 But I don't think there's anything wrong with gradually growing your lab. 434 00:35:48,240 --> 00:35:54,560 As we were talking about before with some of the shortage of the technical experts, 435 00:35:54,560 --> 00:35:56,280 that's another area, right? 436 00:35:56,280 --> 00:36:03,640 Could you potentially, this might be a little naive, but could you potentially teach somebody 437 00:36:03,640 --> 00:36:09,480 to do the lab tech work without necessarily having an advanced degree to meet that kind 438 00:36:09,480 --> 00:36:14,760 of demand and still be effective at what they would be doing? 439 00:36:14,760 --> 00:36:16,720 Well I'm sort of living proof. 440 00:36:16,720 --> 00:36:23,280 So I just showed it at the laboratory and I had a degree in economics. 441 00:36:23,280 --> 00:36:32,600 I principally got the job because like I said, there's a real shortage in people with, it's 442 00:36:32,600 --> 00:36:33,600 interesting. 443 00:36:33,600 --> 00:36:38,440 So you'd think there'd be more, but there is a shortage with people with scientific 444 00:36:38,440 --> 00:36:44,280 degrees that are interested in doing lab work and aren't already snatched up, especially 445 00:36:44,280 --> 00:36:47,880 people with experience. 446 00:36:47,880 --> 00:36:51,560 And then, so it's tough to find people. 447 00:36:51,560 --> 00:36:55,160 So I had a degree in economics. 448 00:36:55,160 --> 00:37:00,640 So I think the lab manager thought, okay, maybe this person can end up doing statistics 449 00:37:00,640 --> 00:37:02,560 at some point. 450 00:37:02,560 --> 00:37:08,520 But it's really just, we just need somebody competent who can just read the, what's called 451 00:37:08,520 --> 00:37:11,720 the standard operating procedures, the SOPs. 452 00:37:11,720 --> 00:37:17,800 So we just need someone who can just read those and just go through the steps because 453 00:37:17,800 --> 00:37:20,200 it's not the trickiest thing in the world. 454 00:37:20,200 --> 00:37:32,320 And honestly, I think lab work's real fun and it's fun work because you basically have 455 00:37:32,320 --> 00:37:34,080 got your steps there. 456 00:37:34,080 --> 00:37:36,720 And so you can just go through the steps. 457 00:37:36,720 --> 00:37:42,400 And yes, it does help to like know a lot about the reagents and the chemicals, but you have 458 00:37:42,400 --> 00:37:44,000 all the steps right in front of you. 459 00:37:44,000 --> 00:37:55,000 So it says pretty clearly like, okay, you want to pipette some, pipette like one milliliter 460 00:37:55,000 --> 00:38:01,160 of methanol into this vial and then put some cannabis in there. 461 00:38:01,160 --> 00:38:07,600 And then go through these steps. 462 00:38:07,600 --> 00:38:16,760 So it's pretty straightforward, but you need someone preferably with a degree in science 463 00:38:16,760 --> 00:38:24,600 who can go through and follow the steps and just be reliable and all that. 464 00:38:24,600 --> 00:38:34,760 For whatever reason, it's hard to, yeah, it's hard to find lab analysts, I think. 465 00:38:34,760 --> 00:38:35,760 Yeah. 466 00:38:35,760 --> 00:38:36,760 That's interesting. 467 00:38:36,760 --> 00:38:37,760 Yeah. 468 00:38:37,760 --> 00:38:44,240 So I could see some bright entrepreneur that's working in a lab that could go, I could boil 469 00:38:44,240 --> 00:38:49,360 this down to like you're talking about some of these lab analyst type steps, like recipes 470 00:38:49,360 --> 00:38:53,120 or whatever, and then start training these people to meet that demand. 471 00:38:53,120 --> 00:38:56,560 That is an opportunity there too, it seems like. 472 00:38:56,560 --> 00:38:59,000 Oh, well, like definitely. 473 00:38:59,000 --> 00:39:06,400 And like Heather, you may seem like you could do something like this, where it's just like, 474 00:39:06,400 --> 00:39:13,080 you just basically kind of give people instructions on, okay, this is how you pipette, you hold 475 00:39:13,080 --> 00:39:15,080 the pipette straight. 476 00:39:15,080 --> 00:39:20,560 You don't tilt it or you're going to get a wonky measurement. 477 00:39:20,560 --> 00:39:24,720 So there's just little things like that. 478 00:39:24,720 --> 00:39:29,560 And then once you kind of have given your people instructions, then they're off to the 479 00:39:29,560 --> 00:39:31,560 races. 480 00:39:31,560 --> 00:39:36,360 I think there's a, like you said, there's a, not even in the lab space, but everywhere, 481 00:39:36,360 --> 00:39:38,880 but it's just essentially just a video tutorial. 482 00:39:38,880 --> 00:39:44,640 So if you're an expert, you would basically maybe have like a video tutorial on how you 483 00:39:44,640 --> 00:39:58,320 could test microbial or test microbes in cannabis, or this is how you do an extraction for pesticides. 484 00:39:58,320 --> 00:40:12,800 This is how you clean the filter on your HPLC, or this is how you do inside the HPLC, they 485 00:40:12,800 --> 00:40:15,520 have what's called the column. 486 00:40:15,520 --> 00:40:19,880 And sometimes you have to change them and like put in a new one. 487 00:40:19,880 --> 00:40:28,360 And it's a really intimidating thing for someone who may not have like a background in science, 488 00:40:28,360 --> 00:40:32,600 because it's basically like, you're going to have to kind of take apart the instrument, 489 00:40:32,600 --> 00:40:34,880 maybe like unscrew some things. 490 00:40:34,880 --> 00:40:42,440 And it's a little involved, but it's one of those things where if you learn, it's real 491 00:40:42,440 --> 00:40:43,440 valuable. 492 00:40:43,440 --> 00:40:52,200 So I think there, like you said, there's value to be added where you could just have simple 493 00:40:52,200 --> 00:40:58,440 video instructions on how you can go about testing and maintaining the lab. 494 00:40:58,440 --> 00:40:59,440 Yeah. 495 00:40:59,440 --> 00:41:02,120 So I've got a background in aviation. 496 00:41:02,120 --> 00:41:03,800 I've done some flying. 497 00:41:03,800 --> 00:41:12,280 And then when I was in the Air Force as a missile launch officer, working with the ICBMs, 498 00:41:12,280 --> 00:41:15,520 that was all checklist driven. 499 00:41:15,520 --> 00:41:22,880 And these checklists were extensive and flying is a checklist discipline activity as well. 500 00:41:22,880 --> 00:41:24,480 You can write a checklist for just about anything. 501 00:41:24,480 --> 00:41:29,080 I know a lot of surgeons, a lot of doctors use checklists in the OR and stuff. 502 00:41:29,080 --> 00:41:33,040 So I mean, it's just people rely on them heavily. 503 00:41:33,040 --> 00:41:37,480 And it sounds like this could be one of those spaces that could rely on. 504 00:41:37,480 --> 00:41:43,200 Of course, you definitely given the opportunity, you want somebody who can run a checklist, 505 00:41:43,200 --> 00:41:48,280 but they also have the context of understanding of why they're doing what they're doing because 506 00:41:48,280 --> 00:41:50,640 that helps. 507 00:41:50,640 --> 00:41:57,040 But it seems at the bare minimum for a lot of these lab tech roles, some of the work 508 00:41:57,040 --> 00:42:03,600 could be boiled down to checklists and for the more simplistic things. 509 00:42:03,600 --> 00:42:09,760 Definitely, and this actually kind of brings us full circle. 510 00:42:09,760 --> 00:42:15,960 So if the method development manager is spending their time doing data entry, they're not going 511 00:42:15,960 --> 00:42:23,480 to have very much time to train the new chemist or the new lab analyst. 512 00:42:23,480 --> 00:42:29,320 And that's a critical, critical part of the whole operation. 513 00:42:29,320 --> 00:42:38,880 You would be surprised that when things are running well, the senior chemists and the 514 00:42:38,880 --> 00:42:45,120 method development managers, they may spend a lot of their time essentially mentoring 515 00:42:45,120 --> 00:42:56,720 or explaining things to the junior chemists or the lab analysts. 516 00:42:56,720 --> 00:43:04,080 So even the method development managers, everybody can learn from each other. 517 00:43:04,080 --> 00:43:10,600 The micro team can learn from the gas chromatography team. 518 00:43:10,600 --> 00:43:15,240 So that's interesting. 519 00:43:15,240 --> 00:43:18,800 So thank you for sharing all this. 520 00:43:18,800 --> 00:43:22,480 I actually now have a general picture in my mind based around the instruments that you've 521 00:43:22,480 --> 00:43:31,400 been talking about, like the HPLC and the GC, was it LC, SMS, the bio, is like a bio 522 00:43:31,400 --> 00:43:35,400 lab team for looking at microbes and that sort of thing. 523 00:43:35,400 --> 00:43:39,960 So just by virtue of you just talking about these different spaces, I kind of started 524 00:43:39,960 --> 00:43:45,760 to get a picture in my mind of the different teams or functional groups within the lab. 525 00:43:45,760 --> 00:43:47,960 So that's pretty cool. 526 00:43:47,960 --> 00:43:48,960 Exactly. 527 00:43:48,960 --> 00:43:55,240 I mean, there are like small labs where people wear a lot of different hats. 528 00:43:55,240 --> 00:44:02,680 It's sort of when your lab grows, these become like different departments because they're 529 00:44:02,680 --> 00:44:12,400 real specialized and you want almost like someone with a degree in microbiology or biology 530 00:44:12,400 --> 00:44:16,840 in your testing for microbes. 531 00:44:16,840 --> 00:44:20,000 And there's different types of chemistry. 532 00:44:20,000 --> 00:44:27,800 So certain chemists are really good at analyzing heavy metals. 533 00:44:27,800 --> 00:44:34,880 And then others are just really good at using like the liquid chromatography for cannabinoids. 534 00:44:34,880 --> 00:44:39,240 And then so everyone has their own sort of expertise. 535 00:44:39,240 --> 00:44:46,320 So yeah, I was just popped into my head, but you know, when you take your car into a dealership 536 00:44:46,320 --> 00:44:51,320 to get worked on, right, you've got all these different folks that know different aspects 537 00:44:51,320 --> 00:44:55,080 of the car's function and they know a lot about troubleshooting and fixing issues and 538 00:44:55,080 --> 00:44:56,680 things like that. 539 00:44:56,680 --> 00:45:01,200 But sometimes you just need to take it to the jiffy lube, right, and have a certain 540 00:45:01,200 --> 00:45:05,880 thing done and it's done quickly and it's done well over and over again. 541 00:45:05,880 --> 00:45:12,920 And it just, it seems like they're, I don't know, my prediction of, you know, in the future 542 00:45:12,920 --> 00:45:18,440 that I imagine that somebody somewhere is going to see this opportunity and gobble it 543 00:45:18,440 --> 00:45:23,280 up and turn it into a standardized, as much as you can, obviously. 544 00:45:23,280 --> 00:45:29,360 You're talking about the method develop manager and there's definitely an art and a science 545 00:45:29,360 --> 00:45:35,520 to this, but I can see somebody optimizing this opportunity at some point. 546 00:45:35,520 --> 00:45:36,520 They're trying. 547 00:45:36,520 --> 00:45:37,520 Are they? 548 00:45:37,520 --> 00:45:44,640 And on all the board, and so it's, you know, if they can do it well and just test things 549 00:45:44,640 --> 00:45:50,920 well and cheaply, then I mean, let's do that because right now, like I said, there's a 550 00:45:50,920 --> 00:45:53,160 ton of improvement that needs to be made. 551 00:45:53,160 --> 00:45:59,200 So we'll see. 552 00:45:59,200 --> 00:46:01,160 Like I said, there are definitely people trying. 553 00:46:01,160 --> 00:46:09,200 So there are like big laboratories, multi-state laboratories, but I would love to see how 554 00:46:09,200 --> 00:46:10,560 the sausage is made. 555 00:46:10,560 --> 00:46:15,840 I would love to see how organized their data is. 556 00:46:15,840 --> 00:46:18,400 Yeah, good stuff. 557 00:46:18,400 --> 00:46:19,400 Good conversation. 558 00:46:19,400 --> 00:46:20,400 Definitely. 559 00:46:20,400 --> 00:46:23,760 I'm not dissing them. 560 00:46:23,760 --> 00:46:30,480 I'm sure, I think there's some phenomenal labs out there. 561 00:46:30,480 --> 00:46:38,080 Any multi-state laboratory probably has a grid system to keep things organized. 562 00:46:38,080 --> 00:46:40,160 Yeah. 563 00:46:40,160 --> 00:46:47,200 But I don't want to keep, I keep asking questions. 564 00:46:47,200 --> 00:46:51,920 I don't want to keep Tony in the conversation, so I'll be quiet. 565 00:46:51,920 --> 00:46:54,800 Oh no, it's good to get some feedback. 566 00:46:54,800 --> 00:47:03,720 Yeah, I guess we were going to talk about like sales data and things today, but myself 567 00:47:03,720 --> 00:47:08,960 and now Charles, we're just going to be in the weeds with this, at least for the next 568 00:47:08,960 --> 00:47:09,960 week or so. 569 00:47:09,960 --> 00:47:10,960 Yeah. 570 00:47:10,960 --> 00:47:18,360 But we're just going to try to make our contribution here because like you said, there's a whole 571 00:47:18,360 --> 00:47:19,840 lot that could be done. 572 00:47:19,840 --> 00:47:25,720 And like, I'm not going to pretend I'm going to, like, so you, like kind of like you were 573 00:47:25,720 --> 00:47:33,800 talking about with, or maybe we were talking about with processors and cultivators where 574 00:47:33,800 --> 00:47:40,480 they can start automating this, like assembly line style. 575 00:47:40,480 --> 00:47:45,320 People would love to do that in the lab space because just like, oh yeah, just put the sample 576 00:47:45,320 --> 00:47:50,440 here and it'll grind it up and put it on the HPLC and test it. 577 00:47:50,440 --> 00:47:54,240 And all you have to do is sign the certificate and send it out. 578 00:47:54,240 --> 00:48:01,600 But that's like, like we said at the beginning in the ideal world, that's how things would 579 00:48:01,600 --> 00:48:02,600 work. 580 00:48:02,600 --> 00:48:10,600 And it's just, just real smooth automated, but in practice, there's a lot of moving pieces. 581 00:48:10,600 --> 00:48:11,600 Yeah. 582 00:48:11,600 --> 00:48:12,600 Yeah. 583 00:48:12,600 --> 00:48:13,600 Well, you gave me a taste of that today. 584 00:48:13,600 --> 00:48:15,600 There's tons. 585 00:48:15,600 --> 00:48:16,600 Yeah. 586 00:48:16,600 --> 00:48:20,200 And it's like, and that role standardized. 587 00:48:20,200 --> 00:48:24,240 So there's different vendors. 588 00:48:24,240 --> 00:48:30,960 You'll learn quickly when you talk with scientists that they all have real strong opinions about 589 00:48:30,960 --> 00:48:33,000 how things should be done. 590 00:48:33,000 --> 00:48:42,720 So, so it's fun when you get a couple of scientists in a room. 591 00:48:42,720 --> 00:48:43,720 Yeah. 592 00:48:43,720 --> 00:48:48,600 It's been a good conversation. 593 00:48:48,600 --> 00:49:00,800 So I guess we've got 10 minutes left. 594 00:49:00,800 --> 00:49:09,280 Do you want to just look at a forecasted some Colorado data real quick or just call it a 595 00:49:09,280 --> 00:49:12,280 day or it's up to you guys. 596 00:49:12,280 --> 00:49:15,800 I know that you mentioned last week that you might be showing something. 597 00:49:15,800 --> 00:49:18,080 So if you have too much trouble, I'd like to see it. 598 00:49:18,080 --> 00:49:19,080 Oh yeah, sure. 599 00:49:19,080 --> 00:49:20,600 We've got 10 minutes left. 600 00:49:20,600 --> 00:49:34,200 So we can run through, I do some data analysis real quick. 601 00:49:34,200 --> 00:49:43,400 So Colorado, like we were saying, does a fairly good job at publishing their data and better 602 00:49:43,400 --> 00:49:46,240 than I thought, unless it was updated recently. 603 00:49:46,240 --> 00:49:53,600 So at first I thought you just had to get monthly reports, which are useful because 604 00:49:53,600 --> 00:49:56,640 you have sales by county. 605 00:49:56,640 --> 00:50:05,480 And I sort of had mentioned this to Paul, but an interesting thing you could do here 606 00:50:05,480 --> 00:50:20,760 is basically run a regression of total sales in each county on the county's medium income 607 00:50:20,760 --> 00:50:28,360 and the population, because I realized Paul that you want to control for population because 608 00:50:28,360 --> 00:50:35,920 obviously places with the high population are going to have higher sales. 609 00:50:35,920 --> 00:50:47,800 But it would be interesting to see if the medium income affects total sales in each 610 00:50:47,800 --> 00:50:48,800 county. 611 00:50:48,800 --> 00:50:54,760 I don't think that's true by the population. 612 00:50:54,760 --> 00:51:02,080 The article I read about the counties along the Oregon-Idaho border, they actually have 613 00:51:02,080 --> 00:51:09,760 a higher per capita sales than in Multnomah County, which is the Portland area, because 614 00:51:09,760 --> 00:51:14,280 you have an influx of people coming from over the border. 615 00:51:14,280 --> 00:51:18,800 So you basically would need a... 616 00:51:18,800 --> 00:51:22,160 I would have to think about how you would interpret the results. 617 00:51:22,160 --> 00:51:25,920 You always have to think about how you interpret the results when you start tossing a lot of 618 00:51:25,920 --> 00:51:34,280 variables, but you could basically add a control, like a zero or one, if the county is a border 619 00:51:34,280 --> 00:51:38,280 county. 620 00:51:38,280 --> 00:51:51,720 So that would maybe control for the cross state sales, theoretically. 621 00:51:51,720 --> 00:51:56,160 But just to keep this moving quickly, that is an analysis you could do. 622 00:51:56,160 --> 00:52:08,400 For now, we're just grabbing these totals, which are actually conveniently in this historical 623 00:52:08,400 --> 00:52:10,400 report. 624 00:52:10,400 --> 00:52:25,280 So I already had some Colorado data, and so I just sort of depended this on. 625 00:52:25,280 --> 00:52:41,360 So I really need to flesh this data out, because Colorado also does these annual reports. 626 00:52:41,360 --> 00:52:48,840 So I think they're quarterly, but they don't publish them that frequently. 627 00:52:48,840 --> 00:52:59,680 But if you look at one of these, you have some amazing data points here. 628 00:52:59,680 --> 00:53:08,800 So you have the new businesses. 629 00:53:08,800 --> 00:53:16,960 I think you have the expired licenses, and then the ones renewed. 630 00:53:16,960 --> 00:53:23,880 So this may have been where I was calculating the entries and exits from. 631 00:53:23,880 --> 00:53:34,240 But long story short is, I haven't actually parsed this data yet, or at least not in an 632 00:53:34,240 --> 00:53:36,800 automated way. 633 00:53:36,800 --> 00:53:45,880 So there's fruit here to be had, but it may be high-hanging fruit. 634 00:53:45,880 --> 00:53:53,040 But nonetheless, just wanted to show you this data, because you can get some good plants 635 00:53:53,040 --> 00:53:54,040 here. 636 00:53:54,040 --> 00:53:57,040 That was cool. 637 00:53:57,040 --> 00:54:06,840 But we're just going to just look at the totals, total sales, just for simplicity's sake for 638 00:54:06,840 --> 00:54:07,840 today. 639 00:54:07,840 --> 00:54:15,320 And in fact, we had looked at this data back in March or so, and I really should have done 640 00:54:15,320 --> 00:54:16,320 the forecast then. 641 00:54:16,320 --> 00:54:23,440 And I thought I was planning on it, but somehow I didn't get around to it. 642 00:54:23,440 --> 00:54:24,800 It doesn't look like. 643 00:54:24,800 --> 00:54:30,160 So now is better than never is the Python philosophy. 644 00:54:30,160 --> 00:54:38,800 So we're just going to do some quick forecasts of this Colorado data. 645 00:54:38,800 --> 00:54:49,760 So we'll just kind of walk through this. 646 00:54:49,760 --> 00:54:57,080 So just real quick to just show you what the data looks like. 647 00:54:57,080 --> 00:55:20,960 So as we see, Colorado has quite impressive sales. 648 00:55:20,960 --> 00:55:33,400 Just to look at just the last handful of observations. 649 00:55:33,400 --> 00:55:51,760 In April of this year, they did 200 million in cannabis sales in Colorado, and it's just 650 00:55:51,760 --> 00:55:56,120 hard to really comprehend the scale of that. 651 00:55:56,120 --> 00:56:03,600 I mean, it may make sense, I guess, based on the population in Colorado, but to me, 652 00:56:03,600 --> 00:56:07,240 it seems just like a staggering amount of revenue. 653 00:56:07,240 --> 00:56:16,680 And it seems like there's some seasonality in the sales here because you've got dips 654 00:56:16,680 --> 00:56:18,960 in wintertime and peaks in summertime. 655 00:56:18,960 --> 00:56:24,440 Do you think that's due to just harvesting times or for outdoor crops, or is there something 656 00:56:24,440 --> 00:56:25,440 else going on? 657 00:56:25,440 --> 00:56:27,840 It's related to the economy. 658 00:56:27,840 --> 00:56:38,160 So as we were talking about earlier, there's some studies, and I'm going to add them to 659 00:56:38,160 --> 00:56:44,560 the repository after this, but just to back up this claim, so I'll correct myself if I'm 660 00:56:44,560 --> 00:56:45,560 wrong. 661 00:56:45,560 --> 00:56:54,080 Unless I'm wrong, I believe studies show that as your income rises on average, your cannabis 662 00:56:54,080 --> 00:56:56,680 consumption decreases. 663 00:56:56,680 --> 00:57:10,320 So there's an interesting play in cannabis sales related to economic growth, really. 664 00:57:10,320 --> 00:57:18,440 So basically, when the economy is doing well, people's income increases on average. 665 00:57:18,440 --> 00:57:23,580 And so there's counteracting effects. 666 00:57:23,580 --> 00:57:31,720 So there's an income effect where people generally have more income, and when you have more income, 667 00:57:31,720 --> 00:57:34,240 you just buy more of everything. 668 00:57:34,240 --> 00:57:38,320 So when you have more income, you'll buy more cannabis. 669 00:57:38,320 --> 00:57:48,320 But then there's a substitution effect where as people have higher incomes, they spend 670 00:57:48,320 --> 00:57:57,160 a relatively lower portion of their income on cannabis, and they spend it on other things. 671 00:57:57,160 --> 00:58:01,200 I'm not sure why this is, maybe. 672 00:58:01,200 --> 00:58:12,440 I don't know why, but this is just what people have observed. 673 00:58:12,440 --> 00:58:15,280 Which effect dominates? 674 00:58:15,280 --> 00:58:21,440 It actually kind of looks like the income effect dominates where when the... 675 00:58:21,440 --> 00:58:29,840 Well, the pandemic was a strange time, but generally I feel that cannabis sales spike 676 00:58:29,840 --> 00:58:38,520 when economic growth spikes and goes down with economic growth, but they're not perfectly 677 00:58:38,520 --> 00:58:41,880 correlated by any means. 678 00:58:41,880 --> 00:58:44,640 That's cool. 679 00:58:44,640 --> 00:58:49,000 Because Washington State, do you know if Washington State has a lot of summary stats like this, 680 00:58:49,000 --> 00:58:52,840 like Colorado does? 681 00:58:52,840 --> 00:58:56,200 I'll need to check. 682 00:58:56,200 --> 00:58:57,880 I could do some digging around and checking. 683 00:58:57,880 --> 00:59:01,240 I just would wonder off the top of your head if you knew. 684 00:59:01,240 --> 00:59:08,640 I mean, for some reason, I just don't think they do, but there are some services, more 685 00:59:08,640 --> 00:59:15,840 than a few third parties that do aggregate the total sales. 686 00:59:15,840 --> 00:59:19,600 For some reason, I don't think the state does. 687 00:59:19,600 --> 00:59:26,600 I think they were, but under their current traceability system, I don't think they're 688 00:59:26,600 --> 00:59:28,640 doing summary statistics. 689 00:59:28,640 --> 00:59:29,640 Okay. 690 00:59:29,640 --> 00:59:33,640 That's cool stuff. 691 00:59:33,640 --> 00:59:43,400 And then just to just wrap up real quick, we can just do a quick forecast just with 692 00:59:43,400 --> 00:59:44,400 ARIMA. 693 00:59:44,400 --> 00:59:52,120 And so I'll get the presentation where we talked about ARIMA uploaded, but basically 694 00:59:52,120 --> 00:59:56,480 just using past observations to predict the future. 695 00:59:56,480 --> 01:00:08,240 And as you can see, our forecasts really aren't that much more informative than really we're 696 01:00:08,240 --> 01:00:10,800 just kind of continuing the trend. 697 01:00:10,800 --> 01:00:19,720 But it's useful because we can kind of get a rough estimate of what the year total may 698 01:00:19,720 --> 01:00:20,720 be. 699 01:00:20,720 --> 01:00:31,960 And so now we can get an early estimate that we're looking at, like 2.5 billion in cannabis 700 01:00:31,960 --> 01:00:33,840 sales in Colorado. 701 01:00:33,840 --> 01:00:34,840 Wow. 702 01:00:34,840 --> 01:00:43,560 And just to give you some perspective, in my hometown, they got a new elementary school, 703 01:00:43,560 --> 01:00:56,800 and I think they spent 2 million on their new elementary school. 704 01:00:56,800 --> 01:01:04,920 So you could, if you had spent all of that on schools, you could buy, or they could find 705 01:01:04,920 --> 01:01:10,480 1200 new elementary schools. 706 01:01:10,480 --> 01:01:14,200 Well, that's total revenue. 707 01:01:14,200 --> 01:01:19,240 So they're not getting all of that in taxes, but they're getting a heck of a lot in taxes. 708 01:01:19,240 --> 01:01:26,800 They're probably getting, I think in Colorado, 25% or so in taxes. 709 01:01:26,800 --> 01:01:33,320 Maybe not that much, but I think in Oklahoma, they're getting 7%. 710 01:01:33,320 --> 01:01:47,640 So you could maybe make between maybe 80 and 300 elementary schools with the amount of 711 01:01:47,640 --> 01:01:51,760 tax revenue that Colorado is bringing in. 712 01:01:51,760 --> 01:01:57,360 So I always kind of like to kind of put things in perspective. 713 01:01:57,360 --> 01:02:06,480 And I thought that was an interesting way to make these huge amounts of money more concrete. 714 01:02:06,480 --> 01:02:13,720 Yeah, it's a good data storytelling technique there. 715 01:02:13,720 --> 01:02:24,560 Exactly because I'm not saying they're necessarily putting their money towards education, but 716 01:02:24,560 --> 01:02:35,040 you know, that's a lot of public good that this was revenue that was getting captured 717 01:02:35,040 --> 01:02:41,280 by, in some cases, not necessarily the best actors. 718 01:02:41,280 --> 01:02:51,920 I mean, I'm not saying everybody in the illegal cannabis markets are bad people, but I'm sure 719 01:02:51,920 --> 01:02:54,080 there are some shady characters there. 720 01:02:54,080 --> 01:03:01,000 So at least now, the people making revenue, they have to file for a license. 721 01:03:01,000 --> 01:03:05,520 They typically get a background check. 722 01:03:05,520 --> 01:03:09,720 So they're at least not criminals. 723 01:03:09,720 --> 01:03:17,800 So anywho, I personally kind of think that's a positive thing, but think of it what you 724 01:03:17,800 --> 01:03:18,800 will. 725 01:03:18,800 --> 01:03:19,800 Sounds good. 726 01:03:19,800 --> 01:03:25,160 But we've run a little extra here. 727 01:03:25,160 --> 01:03:37,040 So thank you for staying tuned and seeing that quick presentation of data. 728 01:03:37,040 --> 01:03:40,000 Thanks for answering all those questions. 729 01:03:40,000 --> 01:03:41,440 I really appreciate it. 730 01:03:41,440 --> 01:03:45,640 And good luck to you and Charles in the week or so coming up with the work that you're 731 01:03:45,640 --> 01:03:46,640 doing. 732 01:03:46,640 --> 01:03:47,640 Definitely. 733 01:03:47,640 --> 01:03:51,040 Thanks for coming, Paul and Heather and Charles. 734 01:03:51,040 --> 01:03:53,360 So thanks for listening. 735 01:03:53,360 --> 01:03:56,200 It's always fun to share. 736 01:03:56,200 --> 01:04:02,080 And then until next week, we'll be making laboratory software and check in next week 737 01:04:02,080 --> 01:04:08,680 and we should have a lot to tell about how things went. 738 01:04:08,680 --> 01:04:11,760 And hopefully we've got some software up and running. 739 01:04:11,760 --> 01:04:12,760 Sounds good. 740 01:04:12,760 --> 01:04:13,760 All right. 741 01:04:13,760 --> 01:04:14,760 Thanks guys. 742 01:04:14,760 --> 01:04:15,760 Take care. 743 01:04:15,760 --> 01:04:16,760 Bye. 744 01:04:16,760 --> 01:04:17,760 Bye.