HH57-05-11-21_mixdown.mp3-from OneDrive

Avery: [00:00:08] Hello and welcome to everyone else who is joining, I see we got a lot of people joining in. Hello, everyone. I am not Harpreet. Despite you might think, Oh, I know a lot of you guys look shocked. I shaved my beard and this is what Harp looks like with the shaved beard. Ok? Just kidding. My name is Avery. I'm one of my friends. And yeah, I got a haircut too. I'm just filling in for Harpreet today. I think he has his sister visiting in town, and so I am just covering for him in the in the happy hour. So if you guys don't know me, my name is Avery. I like Data. I'm excited to to chat with you guys today about Data. We were kind of we had Carlos and Ken. We were talking about the the idea of you. It's been it's been kind of crazy the last two years. And some of you guys, you

Avery: [00:00:58] Know, it's we don't have as much adult

Avery: [00:01:01] Human human conversation as we did. Probably, you know, pre-COVID, just because we're not working in the office as much. And so some of this conversation is good for us to have, even though it's on Zoom. So welcome in. I had a question. I just kind of wanted to start this off today with with a little bit of a prompt about something that I've been thinking about recently. And I would love to get your guys opinion on it and also feel free to add your own comments or your own topics or own questions that you would like to talk about. But I just thought this would get us going and conditions here. And she is one of the members of Data career jump start, which is a project camp that I run and we had a hackathon the last two weeks. We've been doing a hackathon with a company where I took one of my consulting projects and turned it into kind of a crowdsourcing activity. Try to give them some some valuable insights from this data that they have. And we just we just submitted all of our hackathons, our guys, they submitted their submissions to me and I went [00:02:00] with the CEO of this company and kind of looked at it and decided, you know, what do we think was valuable and what was not valuable? But we ran into this question where a lot of

Avery: [00:02:07] People inside and participating in the

Avery: [00:02:08] Hackathon

Avery: [00:02:09] Were like, How do I submit insights?

Avery: [00:02:12] And so I just wanted to open up to you guys what, as you know, data professionals, what is the best way to to present insights because you might have spent, you know, hundreds of lines in Python or R, or you might have written the world's longest SQL query,

Avery: [00:02:27] But how do you take these

Avery: [00:02:29] Insights that you've developed over days or weeks and package them into something that is useful for a business stakeholder? So I'll just open up the floor. If anyone has any thoughts on how to best package Data insights to a business stakeholder

Carlos: [00:02:44] Go, I guess it depends on what they like. But in general, I find in consulting, they like PowerPoints. So what I do is I try to make interactive PowerPoints using like, you know, slideshow and things like that that like embedded in our shiny app inside a PowerPoint format. So this is what the defaults are. If you wiggle it, we get these results. Other people not in consulting tend to like report governments, love reports. So I try to get them out of word documents into again, like tiny markdown, if possible, so they have some wiggle room and things like that. But I guess it depends on them. But I will just note that like, do not hand a business person a a notebook. You don't do that. And I do that. I keep seeing it.

Avery: [00:03:27] Excellent point. Ken, I saw you had your hand raised. Go ahead.

Avery: [00:03:31] Yeah, that's a great insight by Carlos.

Ken: [00:03:34] Like you like different strokes for different folks

Avery: [00:03:37] Like you have to know

Ken: [00:03:38] Your client or who you're talking to, especially in a consulting setting.

Avery: [00:03:42] Something that I think is really valuable

Ken: [00:03:45] Is giving whoever you're presenting to or conveying this information to a little bit of ownership. Right. And that's making like presenting it in a way that it's also their idea. So I find a very effective way to do that now [00:04:00] is with

Avery: [00:04:00] Dashboards a very

Carlos: [00:04:01] Effective way to do that now is with storytelling and doing those sorts of things that is going to vary by each unique situation or presentation that you're giving. But if you have that in

Avery: [00:04:13] Mind of like, how

Carlos: [00:04:14] Do I present this in a way that the other person is bought in? They believe it, and they are part of the decision making and part of the discussion, and they feel like they're included. We're not talking over this person, we're not talking under this person. To me, that's an incredibly

Avery: [00:04:30] Powerful way to convey

Carlos: [00:04:32] Any information, any new insight, any of these things.

Avery: [00:04:35] I like that

Avery: [00:04:36] Because it's it's people like to feel like they did something right. And so I like that aspect of making it, making the stakeholder feel like it is theirs, because after the end of the day, I mean, they're paying for it, right? So they do kind of it is kind of theirs. They are owning it and going back to a little bit about what Carlos said. You know, the the dashboard is a big thing or just giving them some bells and whistles where they can, like move something they can, you know, move something to the left or move to the right and see a little bit of like what happens with the results. So I appreciate that I did notice, Eric, do you want to do you want to

Avery: [00:05:09] Mention what what you typically

Avery: [00:05:11] Do with with these types of reports?

Avery: [00:05:13] Oh yeah, I just said that I print out my Jupyter notebooks and mailed them to my stakeholders. Sometimes I add some crayon comments in the margins so they know what I'm saying. But no, the thing I was actually going to say, though, was that one thing I learned, like they like, drove home like hard core during my master's program. Alex in same program. So she knows is like your bluff, your bottom line up front. Put that on your first slide. Make it like a no, make it big and make it fast and like so that it's the impactful thing because at the end that they probably won't remember anything else, or you won't get them to remember anything else if you don't catch them right off the bat with something that matters to them? Not necessarily. That's just, like, super exciting for you. Anyway, that's all I had to add.

Avery: [00:05:56] What was that bluff? What does it stand for? I've never heard that before.

Avery: [00:06:00] Love [00:06:00] bottom line up front. It is a great north star. Keep it in mind.

Avery: [00:06:05] I taught me a ton. I do like that. So I did see in the chat. Also, we have Mexico here and everyone else who joined looked like we got Matt and

Avery: [00:06:13] Gina and

Avery: [00:06:14] Other Matt. And let's see who else have I said hello to everyone, Alex and Vin and Archer. Welcome. Welcome. Appreciate all of you guys being here. Do any of you guys have any topics or something that does someone want to? Oh, do we have any hands raised? Yes, we do. Monica, hi, I didn't say hi to you. How are you? Hello, good. How are you? Good. What's up? What's on your mind?

Speaker4: [00:06:36] Yeah, all of these have been really great. I just wanted to add it depends on what you are providing to your stakeholders or customers as well, because I have a lot of background in providing analytics and metrics and such.

Avery: [00:06:48] So dashboard is very

Speaker4: [00:06:50] Heavily involved there, whether it be in Tableau or RBI or whatnot. One important thing to add, though, with any onboarding service to provide is to not just present it to the stakeholder and let them run

Avery: [00:07:06] With it, sit down

Speaker4: [00:07:08] With them and really like walk them through the functionality because a lot of times they don't really know how to interact with the dashboard. They don't even know that you can click and ghrelin or anything like that to really sit down with them and show them all of that functionality up front and hopefully cross your fingers that they won't print it out because I have actually seen that happen as well.

Avery: [00:07:30] Always, always seems a little counterproductive, but hey, some people, some people like their printed stuff I have.

Avery: [00:07:36] This is my

Avery: [00:07:37] Calendar, so I can't deal with digital calendars. I have to I have to print mine out. So I guess I'm archaic in some ways. But but thanks for your point, Monika. That's a great point. Eric, did you have your hand raised?

Avery: [00:07:48] Yeah, just one other thing. I would want to go saying that that is like I was I was thinking this week that probably the number one skill Data e skill I'm working on right now [00:08:00] and it's kind of related to delivering final results to stakeholders is like. The ton of communication with stakeholders that happens all the way, like I'm I'm not improving my modeling skills or my skill skills or whatever, nearly as much as I'm getting better at remembering like I need to put in a ticket for that person because there's no way they're going to just remember this magically in three days that I made this request or something, or making sure I'm adequately commenting things and stuff so that, you know, a couple of weeks down the road, people know that they weren't forgotten and they feel like they're being engaged with, and then they're more likely to engage me back. And, you know, just be my friends because I need more friends for that. So whatever. But anyway, yeah, so like stakeholder communication and has just been so huge and I just think I'm just going to be working on that until I die. Probably.

Avery: [00:08:57] Carlos, go ahead.

Carlos: [00:08:58] Yeah. So actually, I don't know, Monica kind of like woke up my brain up until, like more mechanical than I was thinking in government contracting. We actually have like a very specific list of requirements for deliverables, and they'll include like. You know, recorded demo something would call like, oh, and like operations and maintenance, like an O and M guide, you know, like a run book, which is actual interactions we should expect from an interaction. So I've developed documentation that's like, this is the power bi screen. This is this visual. This is the table that feeds that visual. If you click this, you will get this output. For example, see this appendix, this timestamp of this video like I forgot this how in depth some clients like it. So just to double up on what Monica said of like, do not assume that they know that right click drill through is different. The right click drill down because they will call you wondering why the tap didn't change stuff like that. Yeah, just double just doubling down on, Monica said. And also giving you an example on like the actual words [00:10:00] you can Google O&M run book things like that.

Avery: [00:10:03] Excellent point. I had a I had a history teacher who when when he was in high school, he worked for, I think, Disney World or Disneyland, and he played E.R. and I don't know if this is true today, but this is this is what he told us that he's, you know, he's maybe 70 now. He's he was a teenager a long time ago. But he says when he was at Disney, the number one, the number one rule they would tell you was never underestimate the stupidity of an average American. And I think I think that's probably true for for all of us is, you know, never underestimate how little business stakeholders do know, especially on the technical side, especially with some of this, you know, with some of these, you know, fancier dashboards or something they maybe never have seen, it might be. It might be troublesome.

Avery: [00:10:48] They might not know

Avery: [00:10:49] What they're actually doing, and they might be missing out on a lot of opportunities to actually get more insight than than they are right

Avery: [00:10:54] Now. Vin, I'm

Avery: [00:10:55] Curious to get your take. What's your opinion on delivering deliverables or delivering insights to stakeholders?

Speaker5: [00:11:04] I think I mean, I'm going to kind of repeat maybe some different words what's already been said, but I asked people how they get their insights now. And in a lot of cases, when I phrase it that way, like insights, stakeholders are realized. I don't get a whole lot of insights out of my Data, and they're asking me for a bunch of stuff that isn't really insightful. And so I like starting the conversation about how you get stuff delivered now, kind of raising the expectation of what I'm going to be delivering. I mean, obviously, it's dangerous. I better be able to do it. But what I'm asking them for

Avery: [00:11:38] Is not only how do

Speaker5: [00:11:39] You want to see it, but what do you want to see? Because that's that's the more important question. I mean, you're going to have to deliver something at some point,

Avery: [00:11:47] But more times than not. The first thing

Speaker5: [00:11:49] That I

Avery: [00:11:50] This I ran into this

Speaker5: [00:11:51] About six years ago. The first thing I would deliver wasn't really what they wanted. It was what they told me they needed, but it wasn't really what they wanted. And [00:12:00] so I spend a lot more time

Avery: [00:12:02] Now making sure I deliver what,

Speaker5: [00:12:04] You know, stuff that they actually want the way they want it.

Avery: [00:12:07] But more importantly, I spent a lot of

Speaker5: [00:12:09] Time asking them, What do you really care about? Like what's really important to you, especially when I'm starting up with a new client.

Avery: [00:12:15] I spend a ton of

Speaker5: [00:12:16] Time just saying, you know, what do you really care about? I know what you've been kind of trained to ask Data people for,

Avery: [00:12:24] But like if I was better than

Speaker5: [00:12:26] Every other Data person because obviously. What would you

Avery: [00:12:30] Ask me for? You know, if I could

Speaker5: [00:12:31] Wave a magic wand and work with a great team, which is probably what I have behind me, what would you want? And, you know, kind of get a list and then, OK, how do I make that useful to you and kind of get an idea of because, you know, reports are great. Powerpoints are great. A lot of companies do have really rigid formatting and, you know, government agencies as well. But there's always an opportunity. Yeah, you have to fill in the blank, you know, fill in the boxes. But there's always an opportunity on the side to deliver something better

Avery: [00:13:01] And potentially help that

Speaker5: [00:13:02] Organization understand. You know, we could be getting this in a different way. We could be consuming this in a better way.

Avery: [00:13:10] Maybe we can

Speaker5: [00:13:11] Aggregate some of this data because most, most organizations have too much of this thrown at them,

Avery: [00:13:16] Usually from different organizations

Speaker5: [00:13:18] And inside of their own. And so that's the approach I take. I mean, I guess it's because I come into a lot of situations where I have to clean up messes, and so I'm always expecting there to be a mess there for me to clean up. And so I'll ask some of those questions up front before I'll deliver a project to anybody.

Avery: [00:13:34] Yeah, that was that was a great insight. Ken, Ken has a question. Do you want to vocalize it, Ken? Go ahead if you want to.

Avery: [00:13:41] Yeah. So just real quick to

Carlos: [00:13:43] Follow up on that. How do you personally balance or what are some tips on balancing between what people say they want and like the unarticulated needs about

Avery: [00:13:51] What they like, should

Carlos: [00:13:52] Want from a Data

Avery: [00:13:53] Perspective or

Carlos: [00:13:54] What they don't know that they want? I find that very difficult. Like, where do you get started? You send like the dashboard [00:14:00] first and get feedback. Or do you start with the questions? I mean, probably the right answer is that some sort of iteration, but I'd love any framework you have around that. Yeah, I

Speaker5: [00:14:10] I am overconfident,

Avery: [00:14:13] Ignorant, and

Speaker5: [00:14:14] I think I've said this before,

Avery: [00:14:16] I go in with sort of this

Speaker5: [00:14:18] Overconfidence in my own ignorance and

Avery: [00:14:21] I ask really stupid

Speaker5: [00:14:23] Questions

Avery: [00:14:24] Because most

Speaker5: [00:14:25] Times stupid questions haven't been asked for at least a year, sometimes more than that. And stupid questions reveal assumptions. That's how I start getting to what they really need to understand is I start asking dumb questions, and I know that sounds

Avery: [00:14:40] Like a really, you know,

Speaker5: [00:14:42] It sounds like a bad

Avery: [00:14:43] Framework to follow to

Speaker5: [00:14:45] Ask those questions. But what I find is that usually by the third or fourth meeting where I'm asking those types of

Avery: [00:14:51] Questions, everyone's

Speaker5: [00:14:53] Coming to the same realization that

Avery: [00:14:54] I am that, you know,

Speaker5: [00:14:56] We should be asking these more frequently. We should be testing assumptions more often. Maybe we should be educating people that are brought

Avery: [00:15:03] In better

Speaker5: [00:15:04] So that all of this institutional knowledge, you know, gets to them so they can do their job faster than three or four meetings down the road. And so that's how I

Avery: [00:15:12] Get there

Speaker5: [00:15:13] Is I ask a lot of questions that,

Avery: [00:15:16] You know, most people are afraid of

Speaker5: [00:15:18] Asking because they're afraid they're going to feel

Avery: [00:15:20] Stupid.

Speaker5: [00:15:21] And I guess I'm kind of comfortable in that role.

Avery: [00:15:23] I don't I

Speaker5: [00:15:24] Know that doesn't sound great, but I'm sort of comfortable being ignorant

Avery: [00:15:29] And asking

Speaker5: [00:15:30] The questions assertively and then helping everyone realize that it's OK to ask what they don't know. And that's usually where I start getting to assumptions when I ask the stupid questions. And then when other people ask questions that they used to think were stupid. But compared to what I've just said

Avery: [00:15:46] Is, you know, Einstein level,

Speaker5: [00:15:49] They start thinking, Well, that was a dumb question, but my question is not nearly as dumb as that one. So I now have permission to ask my dumb question because it's smarter than that, and he [00:16:00] seemed to get a positive response.

Avery: [00:16:01] So they start

Speaker5: [00:16:03] Asking different kinds of questions

Avery: [00:16:06] And that starts the unarticulated

Speaker5: [00:16:09] Needs, and it's really giving people permission to ask me for crazy stuff. That's sort of the level two. Once they've started asking questions and they've started articulating like, this is the thing I've always been afraid to

Avery: [00:16:20] Ask because

Speaker5: [00:16:22] I'm scared of what the answer may be, or I'm scared of looking stupid or whatever it is. But the next layer is really getting them to ask for crazy stuff, you know? Hey, I've always wanted this. How far? How close can we get to that? And I want that's that's really where I can add

Avery: [00:16:37] A lot of value and

Speaker5: [00:16:38] Teams can have a lot of value is when we get people to start asking for, you know, have you ever been to Mars? Could you get us there? And those types of things where somebody had an idea of, you know, this would provide a lot of value to the team as sort of an efficiency project or this would provide a lot of value to our

Avery: [00:16:57] Customers as

Speaker5: [00:16:58] Maybe a first stab at monetizing models. And they'll start asking me for stuff and I'll say, you know what? Let's look at that and it gives them permission again. You know, ask me for crazy things. It's not crazy. That's what we're here for. That's why that's why we're all sitting around the table together. Ask for crazy things. If it's impossible, I'll say I'm going to look into it anyway. I mean, what do I know? Maybe it's not impossible. And a lot of times people my age and older, we have a tendency to go on and marvel a little too much. And so I've opened my mind back up and let me look at that. I don't know. Yeah, probably tried that before, but you know, things move quick. Let me take a look at that and it's like a permission thing and you get layers of permission to the point where they start asking you for stuff that really would be valuable for them.

Avery: [00:17:45] Well, that was a lot. Gina, I see you have your hand raised, what do you have to comment on this?

Speaker4: [00:17:50] Yeah, I kind of wanted to loop back

Avery: [00:17:53] On some of the

Speaker4: [00:17:54] Earlier comments as well around, you know, like. It [00:18:00] let's not confuse stupid, you know? Never underestimate the stupidity of the American public, whatever with, shall I say, busy or just the your stakeholders have a lot going on if they're on the business side. They don't have time to learn Tableau. And it may be frustrating to some of us that some of them are

Avery: [00:18:26] Not even, you know, kind of fluent in excel. But that's the reality if you want to be effective.

Speaker4: [00:18:35] Realize that you've got to figure out the best way to present it to your stakeholders. And, you know, ask them, ask them what's useful for them. You know, this is a challenge for everybody. It's a challenge for me. You know, if you put up a chart, that's anything more

Avery: [00:18:53] Than really a line

Speaker4: [00:18:54] Graph or a

Avery: [00:18:55] Simple bar graph, be ready

Speaker4: [00:18:57] To walk them through it succinctly. And, you know, and clearly so that they can actually grasp what it is you're trying to tell them. So I also think of it like this. So like, you know, the kids today, right? Kids look at us older folks, and depending on how old you are, that could be any age, right? I mean, when I was in college, I thought twenty two year olds were old and I'm a lot older than that. Now they are all kids, and the older you get, the more everyone kind of looks alike in terms of age. And you know, I've been thinking about this. It's like at what point when we get older do we seem to turn into old fuddy duddies, especially the way technology just moves ever faster? Well, the bottom line is, as we get to be adults, we've got jobs. We've, you know, maybe some people, some of us have kids, some of us have, you know, needy cats running around. I mean, there's all, you're running a household. You don't have time to learn all of this stuff. That's just the way it is. And so I always like to. [00:20:00] You know, remind myself of this, even when I've worked with people, I mean, we had a Salesforce implementation in my last job, and it was it was not perfect. It was kind of clunky because the head person who wanted this, I think, maybe thought that more could be done with it in a streamlined fashion that could be done. And there were people in my group who

Avery: [00:20:24] Were just like, I'm not going to do this like, I mean,

Speaker4: [00:20:27] If it was more than a few clicks and I and they wanted me to walk them through it and I would walk them through it. And, you know, depending on my relationship with the person, it's kind of like, Look, man, just, you know, just have a little patience. Just step through this a little bit.

Avery: [00:20:45] But the reality

Speaker4: [00:20:46] Is is like, there are people are just extremely impatient with it now. You know, you can't wave a magic wand and have them magically understand it. There are Data, there

Avery: [00:20:56] Is Data that is analyzed

Speaker4: [00:20:57] To help make decisions. But, you know, just kind of put yourself in their shoes if there's any kind of. Frustration around how do I present this in a way that's really effective? And the last little bit, I'll just add in to what Vin said. And I think I mean, Vin said it great, but. Yet I've heard this said in certain consulting contexts, what's the question behind the question? Your client thinks you, they want one thing, but oftentimes there's actually something else underneath it that that they really that that's what they really want and they may not realize it themselves. And so that's where consultants really add value, because if they had value, good ones add value because they understand that there is, at least at the beginning, certainly an iterative process trying to uncover what the what the real value is. Ok, so that's that's my spiel.

Avery: [00:21:53] I love it. It's definitely key to, yeah, to always ask the why behind the why behind the why [00:22:00] and in Vince's consultants were not all bad, I guess. Anyone else have any thoughts about

Avery: [00:22:05] Just presenting and I guess

Avery: [00:22:08] Communicating effectively with stakeholders or to someone else, have a question or a comment they'd like to ask, and we can totally pivot from here. I'm happy either way. So just let me know, what do you guys think? One thing while while we're all thinking collectively one thing, I'll go back to

Avery: [00:22:25] Kind of what Vin

Avery: [00:22:26] Was talking about earlier. So I run a really small consulting firm for data science, right? And I deal with some clients that are probably

Avery: [00:22:35] Pretty small

Avery: [00:22:36] And pretty new into the Data field, and I've recently. So I'll have people reach out to me via LinkedIn or on my website and they'll say, Hey, I have this idea I want to make. And they'll like last for like a price estimate or whatever, right? And I've really found that I am unable to ever give them a price based off of what they send me in an email.

Avery: [00:22:57] Like it never,

Avery: [00:22:58] Ever, ever is like, Oh, that's really clear what you want. And also, it doesn't really allow me to ask, Well, what are you're actually doing?

Avery: [00:23:06] So I'll just share, I'll

Avery: [00:23:07] Just share an anecdote. And maybe you guys, some of you guys will be able to relate. I was working

Avery: [00:23:12] The

Avery: [00:23:12] Ceo of a small company. They were they were a little venture backed, but they're very small. Reached out to me and it's I won't say who, obviously, but it's a wearable device company, OK? Somewhere between

Avery: [00:23:25] Two point five billion and

Avery: [00:23:26] Billion probably lower on

Avery: [00:23:28] The lower end, but they're

Avery: [00:23:30] Making a wearable device for some sort of context, right? And they reached out to me, Hey,

Avery: [00:23:38] We want to make you know, we want to have

Avery: [00:23:40] An algorithm for a wearable device

Avery: [00:23:42] That will count a

Avery: [00:23:43] Certain action. Ok? They want to be able to count action automatically. I said, OK, that's no problem. I have a background in signal processing. That's something I really like, like and enjoy. So I said I was interested. And they said, Can you quote me on this? And I'm like, Well, OK, like, [00:24:00] we got a machine learning algorithm picking out some signals and above the noise, right? And I don't remember exactly if I

Avery: [00:24:07] If we eventually

Avery: [00:24:08] Like, established like an estimate or not. But I get to the point where where I'm like, Well, it's always easier if I like see the data that you have and she sends me like one csb. And and we kind of start the project and I was like, OK, can you send me the rest of the data?

Avery: [00:24:24] And she was like, That's it.

Avery: [00:24:26] It was like one run from like one

Avery: [00:24:29] Person's device, like a random number

Avery: [00:24:32] Of activities and like. And it was like, make a machine learning model. And it was like, Oh, OK, I have to take a step back. I have to explain to the CEO who is not in the tech space at all, right, which is which is totally fine, but I have to explain. Oh, OK. So usually for these things, you have to have at least a couple of runs. If you want to do any sort of machine

Avery: [00:24:52] Learning, you have to have like

Avery: [00:24:53] A lot more runs than just one. So before we actually get started with anything, you know, we're going to have to actually generate some data. And she was she was so confused at first. But I mean, that's what we're dealing with a lot of the time. Like you, you really these people don't necessarily know what it takes to make a machine learning algorithm. All they know is Netflix has one, and Tesla has one, and Facebook has one. So my new business should also

Avery: [00:25:15] Have one, but sometimes they're not

Avery: [00:25:16] Even ready. You're like, Oh, you don't even your steps away from from machine learning. And that can be a hard lesson for these companies to learn. They don't necessarily know what they don't

Avery: [00:25:26] Know, right? So anyone else had

Avery: [00:25:28] Some experience similar where I mean Vince to nice. And I guess then he's like, Well, I don't think that's possible, but I'm going to go ahead and look into it regardless. I'm not as nice. I was like, I can't do that. I'm sorry.

Avery: [00:25:40] But have you ever had like an experience

Avery: [00:25:42] Where you're like, completely on the wrong terms of a stakeholder

Avery: [00:25:46] Or, you know, someone who's asking

Avery: [00:25:47] You to do something?

Carlos: [00:25:49] Yeah, last night I had a project implode because we were given three weeks to solve like [00:26:00] one hundred million dollar hospital investment problem with like 70 rows of Data, and nobody actually trusted the Data, including all the like a thousand dollar an hour experts in the room. And we were like, OK, well, what we'll do is we'll create an application for you that'll kind of turn your opinions into like a model analytic hierarchy process. And then you guys will do the inputs. We'll give you quizzes. We'll do all this like stuff together to kind of quantify your opinions. And there was a few stakeholders that were just like, This is stupid, why are we doing this? And they pretty much said that throughout the whole process. So I don't know why they brought us on. And then at the end, we delivered and they just said, actually, we're going to we're going to all now agree with that guy who's been saying no for three weeks. So we're not going to do this anymore. Like, but really bad burned a lot of bridges. And it was a huge waste of time, and I was there at the very beginning saying, like, yeah, 70 rows of Data none of you trust it isn't going to go well, but you're paying us to do it and you really want it done. So I'll do whatever you need us to. You know what? That's a lesson learned there.

Avery: [00:27:09] Yeah, and it sounds like an unfavorable outcome, right? But at the same time, I think ending the project after three weeks could be viewed as a actually good outcome rather than ending the project after 30 weeks, right? I think I've been part of an organization. So when I worked for Exxon Mobil and I was a data scientist for Exxon Mobil, you know, COVID hit right and all of a

Avery: [00:27:31] Sudden, no one's

Avery: [00:27:32] Driving anymore. There was the Russia Saudi Arabia price war and oil driving oil down. All of a sudden, no one's traveling and oil's already really cheap. So Exxon is making zero dollars, so they have a decision to make right. They have to, you know, they basically laid off or

Avery: [00:27:51] I'm not going into the details

Avery: [00:27:52] Of that, but basically they had some

Avery: [00:27:53] Layoffs, layoffs, right?

Avery: [00:27:55] And we've been doing projects in our data science division that like had [00:28:00] gone on for like a year or two years. And I was like, Guys, this isn't going to pan out. Like, Why are we still working on this? So sometimes. Sometimes it is good. Hey, you know what? This was a terrible idea. This sucked. It's been three weeks. Let's just kill it right

Avery: [00:28:13] Now, and that's way

Avery: [00:28:14] Better than three months. And that's way better than three years. Avery still under

Avery: [00:28:17] Nda?

Avery: [00:28:18] No. But I have some articles coming out, and I just don't Exxon to ever sue me, ever. So I'm just terrified of Exxon and Big Oil in general. Ken? Sorry, go ahead.

Carlos: [00:28:27] You know, I actually had a fairly similar

Avery: [00:28:30] Situation a long time

Carlos: [00:28:32] Ago at a company I interned at. It's a very large Fortune 100 company, and it seemed like a lot of the Data science department was brought on because they saw other really big companies doing this. Essentially, it was it was like a manufacturing business and the project that I was working on, I had to build a model to predict if they should completely take apart an engine

Avery: [00:28:55] Or they should only

Carlos: [00:28:57] Like partially take it apart. And what I came to realize, you know, like this project was framed by my Data science manager, by this whole team,

Avery: [00:29:05] But they didn't look

Carlos: [00:29:06] Into the Data at all, and what we found is the only there was no ground truth related to this. The only information we had about if an engine should be dismantled or not was if it was dismantled by

Avery: [00:29:19] Someone's judgment previously.

Carlos: [00:29:21] So the best possible model I could build would be to recreate the decision making process of the people who were doing it before. There was no ground truth around. If this should be evaluated, if they should actually be taken down, if there was something wrong with the engines and the person that they were paying to do this was, it was just one person who was doing this consistently. He was super against the project. There were so many forces working against it that it was effectively destined to fail and not necessarily a good use case for data science.

Avery: [00:29:54] But it took forever like months for me to

Carlos: [00:29:58] Explain why the Data was [00:30:00] a problem here and why I couldn't build a model on it. And I think that that's a little bit of a cultural thing. I think it is a little bit our job as data scientist to do this reeducation. Right. Because if we're working for this company for a long time, whether we're an employee or a long term consultant or whatever it is like, we don't want them making the

Avery: [00:30:18] Same mistakes in the future. You know,

Carlos: [00:30:21] Like. We want the best for the organization like that is part of what we do and honestly in sports, I look at that as like a large portion of my job is how do I explain Data and like create Data literacy in an organization so that they can do better next time? Or like create this understanding because the projects build on themselves, right? If you're creating this trust, if you're creating this understanding over time, that creates momentum, and that means you don't have

Avery: [00:30:51] To explain it and maybe only have

Carlos: [00:30:53] To explain it, like reiterate it two or three times rather than five times, then the next one and then maybe one time and then they'll eventually get it. So I'm I'm like a big believer that to me, that was one of the best learning lessons that I could have is that, yeah, I was a grad student. Yeah, I was an intern, but I still had to teach. I still had to explain. I still had room to do that in the role that I was in, and I'm really grateful for that.

Avery: [00:31:20] It's a good it's a good skill to have. Yeah, that ground truth, like the the doing supervised learning without that target variable is not an easy thing to do. I have I haven't really quite learned how to how to do that yet. You kind of need some sort of ground truth. I did want to open it up to anyone who maybe hasn't commented yet. I don't know if Khadija or Makiko or Matt or Eric Russell, I know, joined us kind of late here, if any of you guys want to mention something. Feel free to comment on this at all. Any of you guys got anything you guys want to mention?

Speaker4: [00:31:55] I was curious to hear what what people are doing for side projects or for [00:32:00] upskilling right now because, yeah, no, I'm just like looking for ideas for how to kind of continue growing as an engineer. Besides doing like personal

Avery: [00:32:10] Projects and books, but also

Speaker4: [00:32:12] It'd be kind of cool to hear what people are up to.

Avery: [00:32:14] I got I got one today. I'll just I'll be brief. So like I said, I hosted this hackathon over the last two weeks where I don't know how many people actually participated, but it was kind of kind of my my group of students that I've been working with over the last three months, plus an additional group. So we had we had like overall, a pool of over 100 people. And I needed to make certificates for everyone like customized certificates with their name on it, right? And I was like, How do you do that? So I did some Googling and I found a nice little, you know, pill whatever package from Python and was able to like, make a generic, you know, P&G for the actual certificate and then superimpose all of the all of the names on there. So basically, I didn't have to go in like some sort of graphic design or PowerPoint and copy and paste names for like hours. So I'd rather do it in like 10 seconds. So not like data science, but some automation in Python. That was a lot of fun.

Avery: [00:33:17] Yeah, actually. Carlos, do you want to

Avery: [00:33:19] Talk a little bit about what you've been working up to in the Charlie Dao? It's been it's been fun to watch.

Carlos: [00:33:25] Yes, sir. So Charlie does a collective of Data scientists, software engineers, product people and crypto with three people that 40 members so far. The idea is that if you get a bunch of really smart people, put them in a disk or together, ask them to donate two to five hours a month. We think there's going to be some magic that's going to happen. So so far, we're working on a few different projects. Really, the two of them are my projects, but the idea is that anyone in the collective can just propose a project. And if you want to join the project that is, make a channel and they do it. We only ask that if you monetize [00:34:00] you, exit and spin out before you monetize to keep the collective out of all that nonsense. A lot of money. So we're working on some smart contract stuff, which is solidity, not really this chat space, but the other one we're working on is our NFT analytics platform, which is just a fancy way of saying parameter AIs reports on various like NFTs and tokens, with the goal being to help in the business intelligence side. So there's a lot of communities who they raise money, they want to NFT and they earn hundreds of ether, which is four or five hundred thousand dollars, even a few million dollars. And they have it in their Treasury and they really don't know what they're doing with it. They're often they're really good marketers or really good artists, but they're not necessarily people who understand business and we want to help them do and obviously get paid to do.

Carlos: [00:34:50] It's like help them understand not just the sales activity, the finance boring stuff. I find it boring, but also the like community analytics network analysis who's in their community, who understand the like on chain behavior of the people in the community and then ultimately, like social media, also so integrated like social media analytics, mostly Twitter, but in the future, Discord and Reddit. So you're probably thinking sentiment analysis, things like that. But we're also just thinking of like, how does it all interweave right? Like, what does it mean for someone to join your community in the context of buying when you're in NFTs? What kind of on chain signals are there that more people are interested in community? How do you find out what your committee actually likes and dislikes? And really, the whole business intelligence side of like, how do you grow your business when your business is a community of owners of something? They're really excited for both those projects. And if anyone's interested in joining it, please let me know. The only thing you have to do is ask there is no price or entry or anything. We just want people to ask to join as opposed to putting a link online and having people just pop it in without like committing. We really do want like at least two to five hours a month, if possible, of no time meetings. Answer people's questions regarding providing [00:36:00] resources. Yeah, any questions on that? One person said, You're in Data me.

Speaker4: [00:36:07] What types of skills are you guys looking for? You mentioned business skills

Carlos: [00:36:11] And yeah, so for us, we're focused on anything really like we have no barriers on the skills, like we have a few people who are interns or in school who are just they're interested and they want to learn and they're they're willing to just read Leap. Like if you're willing to read something and share it, that's a skill we're looking for. Whether we summarize, there is marketers, we have people, you know, the collective plurality Data scientists. So a lot of people are Python are people. We have Data engineers too. But really any skill set, especially like marketing, is a big one. We're lacking product is one that we're really low on. We have like Greg and this kind of product, although he's really good. Yeah, I was a marketing and product are two big gaps. Neither of those is coding related.

Avery: [00:37:00] I'll even add that, and I also saw we had a question here from from Russell kind of moving into the NFT space. And I know I've talked to Ken kind of about NFTs previously, too. I mean, it's so crazy the amount of volume that's happening in the NFT space for those who maybe haven't been following, you know, OpenSea, which is basically the biggest market where NFTs are bought and sold, had a month where they did what was

Avery: [00:37:24] It, three billion

Avery: [00:37:25] In revenue or something like that, which is a ton of money. So there's a ton of

Carlos: [00:37:30] Volume, sales volume,

Avery: [00:37:31] Volume. There you go. So that is pretty crazy. And the the amount of so it's a big market, right? We're talking, you know, billions and billions of dollars and there's very little infrastructure there.

Avery: [00:37:45] And if you

Avery: [00:37:46] Can have any sort of advantage or

Avery: [00:37:50] Even just even just right

Avery: [00:37:51] Now, you'd be so surprised how low the bar is for any sort of NFT analytics. One of the leaders is basically like just taking data [00:38:00] from, you know, an API or a database and like displaying it on a Google data studio dashboard. And like like that's getting thousands of views a month. Like it's it's it's a pretty low bar right now in terms of the NFT analytics space. So it is an interesting place. I'm a member of the Charlie Down. It's been fun to interact with some of the people over there as well. I will ask Russell's question Has any of you guys seen the recent use of NFTs for high end whiskey bottles? I have not. Carlos, can anyone else?

Carlos: [00:38:30] Yeah, so just just a quick comment on the broader trend of NFTs and R2bees, which is a fancy acronym for real world assets, is the idea that if you have an NFT, it's an on chain record of a person who owns something. And when you have that, you have the base to do anything you want on top of that base. So like once you have the record of loans that you can pay, what's your address? What's your name? Are you old enough to drink whiskey and you can start doing things like, OK, now off chain? Let's give you a thank you thing of whiskey. More famous examples are unisex if you. There are almost 30 billion, eight hundred thousand or a million dollars, a pair of socks that are real. You've burned your NFT and you get a pair of socks in the mail and people are not burning them because if you don't burn them, then you can sell them for hundreds of thousands of dollars. There's sock, there's clothing. There's all kinds of like real world assets that are tied to end fees. And the big thing you want to remember is like, this is just an on chain proof that you own something and you can do anything you want on top of like proof that someone owns something. Proof of community is like the easiest way to think about it is to answer your question. So.

Avery: [00:39:40] Cool. I'm a I I arrestees are such a weird, crazy place, it's it's it's a it's a very interesting place. And like I said, the bar is pretty low for analytics. So if you are looking for a fun side project, I know I did. I did one about a month or a month and a half ago where I was. I was basically looking to [00:40:00] what the community calls Wales, which is these these wallets or people, I guess, that own a surprisingly large amount of, you know,

Avery: [00:40:09] Either you know, money,

Avery: [00:40:11] Either you

Avery: [00:40:12] Know, crypto or

Avery: [00:40:14] Phd. So I did some analysis on some of these wallets and it was like millions and millions and millions of dollars. So that is a fun space. Any other side projects going on right now that you guys have been working on that you guys have been enjoying? Any anything you guys got going on Eric, Monica, Greg, Matt, Eric and you guys got or maybe Alber. I know Albert hasn't LinkedIn challenge going on right now. Yeah.

Speaker5: [00:40:37] Greg, go ahead. Yeah, so this this week has been like kind of weird for me. I've been pummeled by this whole quantum computing challenge in the the use cases were quite interesting because most of them were about optimization, and I learned quite a bit of things like how to use quantum computing power to optimize financial portfolios. A group of assets minimizing the cost. Maximizing returns. I've learned how to perform molecules shapes for chemical use cases in terms of how electrons rotate around atoms that guarantees certain stability in the molecules, which is pretty cool. And then one of my favorite ones was a use case for electricity, where you'd have like an electrical power attached to your your house or a company attaching a power box in your house and delivering that power to the grid [00:42:00] where you can make money, et cetera. And if you have a fleet of power boxes, how do you optimize which area, which zone to give off electricity to? Knowing that the cycle of recharging discharging will over time deplete the unit's performance in how to optimize this whole system using quantum computing. So it's quite cold. It was painful. I learned every time I touch these things and I touch python and stuff, I know I don't want to be a coder, so kudos to any one of you. With thousands of lines of code, because I will be that business guy just thinking you with whatever I can when I partner with you. So it's been cool and I encourage everybody to know. I encourage everybody to just, you know, just do something for fun and see what you learn. So what I got from this one is probably not much on the coding piece, but more like how you translate a business case into something technical and then, you know, create something that that generates value, right? So it was cool.

Avery: [00:43:20] That's awesome. Thank you for sharing. I know you dabble in so much. You have you have a very good broad understanding of so many, so many different topics. It's very impressive. So I'm scared of quantum. Anyone else had any? I'm guessing no one's had experience with actual quantum computing. Am I right? Then has someone asked you to do an entire project only on quantum computers yet? Or not that impossible yet?

Speaker5: [00:43:45] I've had some experience with it, and I think the crazy thing about

Avery: [00:43:49] Actually doing projects

Speaker5: [00:43:51] That are at some point supposed to be

Avery: [00:43:54] Deployed on a quantum

Speaker5: [00:43:56] Something

Avery: [00:43:57] Because like there's three different types of quantum [00:44:00]

Speaker5: [00:44:00] Computers out there, and no one really agrees completely on which one of them is really a quantum computer versus sort of a quantum computer versus the God. You know, the singularity quantum computer. I mean, it's it's one of those things where the more you do in the quantum space, the more you realize nobody really understands any of this. And companies like IBM and Google and everybody else is saying, Yeah, check it out. It's a quantum computer. I got quantum supremacy and it sounds like, you know that dude in Ant-Man, the the guy that just once he starts talking, he won't start stop talking. Everyone that I've ever interacted with

Avery: [00:44:39] Who's applied

Speaker5: [00:44:41] Like actually doing something with quantum sounds like him. Like once you get them

Avery: [00:44:46] Started, it is an avalanche of

Speaker5: [00:44:49] Information and you just can't like. It's overwhelming until you realize everyone has a

Avery: [00:44:55] Different take on it.

Speaker5: [00:44:57] Everyone's built something a little bit differently, and none of it 100 percent works. And every time you actually start asking in depth questions about how quantum computing

Avery: [00:45:08] Works, even with

Speaker5: [00:45:10] Phds, people that have spent twenty five thirty years in

Avery: [00:45:13] The field of quantum

Speaker5: [00:45:15] Everything they get to a point where they just go, you've got to be, you've got to understand. We don't know. And there's, you know, there is no better example that I think it was about 10 years ago. A company in Australia came up with what they said was a quantum hard drive, and somebody asked the stupid question, You know, where's the Data go? And no joke there

Avery: [00:45:37] Was there were three very smart scientists on the panel who went. Um.

Speaker5: [00:45:43] And that's literally quantum computing summed up like if you wanted to put it in a meme, that's it. Oh man, it's kind of like it's going to turn into like exactly that. Like how many of us using computers really understand what's going on under the hood, right? So it's probably going [00:46:00] to be, you know, let the smart folks build those things under the hood and then folks like me come in and push a couple of buttons and spin up couple qubits and perform my calculations right? So I set my parameters and things like that. So do I need to know much about quantum support vector machine, which is one thing that I used for one of the use cases. And typically, qubits are very scalable, right? So any for every column you have as a feature in the classical support vector machine, you need a qubit for that because you need to know how to translate the classical data into a quantum Data to spin your your qubits and perform these probability checks. And as soon as you understand quantum computing calculations or good at probability checks when you work under uncertainty and if you want like accurate classifications or not accurate classifications if you want classification type of inferences like classical computers can give you, then quantum is not for you, right? So what they do is you have a lot of there's something called variational quantum eigenvectors where you can use a hybrid quantum system with classical system, where the quantum does a probabilistic check and then feeds that those probabilities to a classical system in the classical system runs a cost function and then feeds it back to the parameters, back to the quantum system for optimization.

Speaker5: [00:47:39] So sounds familiar, right? So do we need to know all of this under the hood? I'd rather let the, you know, smart folks deal with that and give me the UI. And let me put my numbers in and then I tell somebody, Look, if you put your money in Tesla and hence you're going to make a lot of money and then you sell your real estate, you know, [00:48:00] those are the things that you know. I think the world will continue to be split by folks who like to tinker under the hood versus, you know, folks who just like to use cool tools and generate value for for the business.

Avery: [00:48:15] Excellent point there, Greg. We did a there was a couple posts on LinkedIn this week. I know I had one about how I don't like elitism in Data science, and I know Ben Taylor had another one where he kind of talked about.

Avery: [00:48:30] That he would like rather take like an

Avery: [00:48:32] Average Data scientist and like some PhD Data scientists, sometimes because they miss the whole value prospect and it's a good debate to have. Sometimes it's like there are people in the world that are really smart that are able to build infrastructure that I can't fathom, you know, and make algorithmic algorithms like am I smart enough to, you know, write a new machine learning algorithm? Probably not. Not really. You know, and I'm grateful that there are people that do that. And at the same time, there's people. We also need people like Greg, who's like, I don't want anything to do with programing. I don't want anything to do with quantum, but I'm going to figure, you know, a way that Amazon can make a heck of a lot of money out of this. Or maybe Greg can make a heck of a lot of money out of this. And and both are valuable. And I think sometimes maybe maybe we're harsh, too harsh to the either side. But that's that's kind of my my closing thought, I think is like, you know, there there are people who are really smart, and maybe that's some of you guys, but some of you guys maybe are

Avery: [00:49:27] Like, Oh, I'll just take what the smart

Avery: [00:49:29] People do and apply it to this specific niche. And that'll bring a lot of value. And I think I think that's I think that's very valuable. So anyone else have any any closing comments or anything to, I guess, set us off on the weekend

Speaker5: [00:49:42] At the end of the day, Avery. You know, you might hear those fancy terms in quantum and stuff. The only advantage people can talk about now in terms of business value is time, right? What can you gain in terms of time? So far, classical computing can do a lot of things that quantum can do. The only [00:50:00] advantage right now, people are kind of debating it does it faster, but does doing it faster really bring value? I'll say it depends depends on the use case because yeah, if it takes you a couple of months to spin up a few molecules to create some drugs versus traditional two years, then yes, it's a it's a quantum advantage. Other than that, there's there's nothing else.

Avery: [00:50:26] Sweet. Well, thank you, guys, thank you guys all so much for being here. I promise you that next week I will grow my beard back and not have a haircut. And I will go buy her instead of instead of Avery. So thanks for sharing with me. I don't. I know I don't make it to a lot of these. My Fridays get pretty busy, but good to see all your faces again. And yeah, fun, fun session. Appreciate all you guys being here. Thank you. Yep. Have a good weekend, everyone. Goodbye.