Welcome to another uh awesome episode of Energy Bites. [laughter] Sorry, we uh we're trying to come up with different superlatives to start off the show. But I've been I've been dumping all of our show transcripts into our language model and I came across a bunch of our episodes in a row cuz they were just sorted by the uh podcast title. And literally the first like 10, it's the exact same. Welcome to another exhilarating or exciting episode of [laughter] So, Bobby and I are trying to iterate cuz it sounds too much like GPT right at the hell. hyperbolic it can be. [snorts] Yeah, but um now we're really excited. We've got uh Jessica Iriarte, is that right to say it? Um from Corva. She's the general manager of completions. Yes. And uh you know, also aka water hammer girl. Promise [laughter] you I'll let everyone know. Um but uh no, I'm super excited. I know we've been trying to coordinate for a while, so thanks for taking some time with us. Yeah, absolutely. Thank you for having me. Calfee [clears throat] in here as well. Don't want to forget him. Forget me, it's fine. [laughter] I'm not important in this uh conversation anyway, so it's I'm I'm excited to have you here just cuz I think a lot of your experience has kind of in some ways we've worked on a lot of the same stuff just in parallel or adjacent to each other. And so I think it's going to be a fun fun conversation. absolutely. [snorts] How uh what are you what are you working on at Corva these days? We'll start at present and work back and kind of see [snorts and laughter] see how that goes. But um but you started fairly recently sort of, right? Yeah. And then you got promoted soon after or so. Yeah. So, congratulations. Thank you. Yes, um right now the general manager for completions and that means basically I'm sort of coordinating across all the different teams at Corva. So, working with sales, marketing, ops, uh product strategy, um basically any every team in like solutions engineers just to coordinate like a direction and plan for vision for completions. Um so, that can mean meeting with customers, doing discovery on what we're thinking ahead for, you know, our AI initiatives or what we need for in the product side. How do we market that? How do we make sure that users are aware of all the new things that are coming out? Um you know, if there are anything on like the outside that I can help with or if we have a new trial with the customers. Basically, kind of the entire life cycle of a completions. That's very cool. Yeah. And then just maybe for people who aren't aware, can you just give a high level of like who Corva is in general? I mean, I again, I think most lot of people listening will know and like obviously we're pretty in tune, but like just in case people are aren't aware. Yeah, we are a software company for oil and gas data, real-time data products. And basically, we like to call ourselves the app store for the oil and gas. So, we have our biggest product lines are drilling and completions where we have in drilling, we have about 100 apps. In completions, we have about 40 apps. And we recently launched newly integrated like geo apps, so geo drilling and geo completion. Basically, when you're able to integrate all of the geo steering interpretations into these data sets and basically, in completions, we're able to take in all the real-time data coming from the field, map it, you know, stage it, label it, and provide real-time metrics. So, a lot of different visualizations for like the operational um monitoring, I guess, of of everything that's going in the field, wireline, frac, pump down. Um and then now we're bringing in all of these integrated metrics, as well. Nice. Yeah, I think y'all were one of the first ones, at least in my memory, which is not very good these days with my children [snorts] and everything else going on, [laughter] but I think y'all were one of the first ones that I noticed in the marketplace that weren't just streaming the data and then doing, you know, what I would call engineering analytics. So, actual engineering problems on the well itself, but they were also doing the operational piece because that is in my opinion, coming from the field as a frac engineer, that was such a low-hanging fruit area for us to optimize. Right, like when I came in the field, you know, gas was probably four or five bucks. Oil, I think, was somewhere between 100 and 120. And so, like, when prices are so high, people don't pay attention to those optimizations cuz it's just get it out of the ground as fast as you can. But then [laughter] 2015 hit and lower for longer happened, and [clears throat] everybody's like, "Oh, well, how do we optimize, you know, the efficiencies of Yeah, now it's the opposite. Right, [laughter] reducing, you know, non-productive time. Yeah, NPT. Yeah. Identifying where the NPT sources are, how do you, you know, optimize those, and it took forever. I mean, operators are finally on board this train, but it was one of those things like from the field where you're just like, "I'm looking outside at six sand trucks parked charging us to merge because no one told them that we were shut down because we screened out." You know, just all these logistical Yeah. kind of uh domino effects. these are massive operations, right? Completions and very complex on the technical and the operational side. Well, they're also logistical nightmares, too. Like, ultimately, that's one of the biggest things that, you know, if you compare oil and gas onshore shale operations to like a factory, it would literally be the equivalent of moving the factory and all of the things you need to make the widget from the factory every single time, every couple weeks, right? Yeah, that's a good analogy. It's And so, it's like, it's very hard for the average person to understand the complexity of this, but like, you have to bring everything. Yeah. You know, if you don't have it or you run out, And then you have so many variables that are all the time evolving, right? Where new tools, new uh yeah, new hardware um that you're trying to test, that you're trying to compare. It's like so much information that is really hard to grasp or to like make decisions in real time. Just saying, "Oh, I mean, obviously people in the field are very experienced and uh can take a lot of that into those decision-making, but when you have more and more, the operation just keeps growing into more wells, now simul fracs, tri simul fracs. Um That's the thing, the scalability is also underrated, right? Like a 5% savings on chemicals on one frac or pad Yeah, absolutely. compounds significantly and you do that 100 times a year, you know, depending on your your schedule is and Yeah, I mean, that's a what? Small or, you know, mid-size operation? yeah, yeah, you start doing it across like an Exxon or you know, I remember what they had like 50 fracs going on when we were at RDS. It's insane, like um Yeah, well, and even like I think, you know, cuz I've, you know, because Corva cut their teeth on the drilling side originally, correct? And then they've moved into completions last probably 5, 6, whatever years. [clears throat] Um but I think I saw recently and I wish I could remember the guy's name. I met him at uh one of the Nate parties, but I he had something and was talking about how like the portion of like a drilling completions AFE like is a heavily weighted towards the completion now. Like I mean, like the drilling rig itself H&P I think made the point like it's a very small piece actually in the big scheme of things of like the overall D&C cost. It's the especially that completion side is a huge cost and then we have a lot of levers that you can pull. I mean, like There's so many variables. Yeah, we talk about it. I mean, there's horsepower, there's fluid, there's sand, there's, you know, chemicals. Like you know, there's all these levers that can be pulled to optimize and then it that it's not even talking about the geology. going to say that's before you get to the rock, which is a whole another chemical Completions really is like the junction of technical and operational like disciplines in the oil industry, right? You have the geo and um drilling feeding into completions and then completions being kind of driven by uh reservoir and production metrics. So we are kind of in the middle of pulling all of these things together and trying to optimize it. It's a good way to put it cuz you're you're right, everything leads up to that and then everything hinges off of that, right? Yeah. But one of the things that I was I love that Corva has done and I hope more people continue to do is like your App Store, this partnership idea, right? Like Bobby and I had have had many discussions both prior and current where we talk about, you know, like if people would just share or focus on what they do really well and let the other people that do what they do really well. And then if you could work together, you can both do your things really well and everybody's happy, right? It is one Yeah, it's definitely one of our strengths. Um we also started leveraging kind of like our platform that we have to deploy our own apps to have customers deploy their own apps in DevCenter. And uh we've also started to deploy our own models. Eventually, we'll allow to customers to bring their own models as well. So, one of our biggest like hits last year was predictive drilling, which is uh actually when I joined Corva, I was leading the data science team working on predictive drilling and then I switched to this general manager of completions role recently. And um it's really fascinating this predictive drilling uh product because basically we we partner with Nabors for it. Um and basically it's sort of like the self-driving rig idea. Not quite yet, but uh getting there on basically having this RAP optimizer that makes predictions in real time and it's able to control the rig and sort of let it take over. Yeah, that'd be really cool. I mean, what Colin was just saying what a week or two ago, he thinks what, probably 5-10 years you can see some rigs and stuff. Yeah. No, I mean, that's that's it's funny because it's [clears throat] like it one, drilling I feel like had a step up on the completion side for 20 plus years because of the EDR, right? We didn't really have that idea in completions. It was just every service provider collects it all and then everybody just throws it to the operator and they get to deal with it. But now, finally, people realize that, "Hey, we can they get are really good at the drilling side and we can't control a rig if we don't own it. So, how are we going to do this even though, you know, the market wants something that can do this." Uh you know, predictive or automated drilling. And so, yeah, to me it's one of those things where it's like the only way you get there is going and buying a drilling company, which isn't going to happen, or partnering. And so, I'm I'm just so so happy. I remember when that announcement came out, and then the uh I can't remember who what frac company it was, but they also had an announcement with another frac provider, and it's like these dominoes are starting to fall, and things are actually finally starting to to happen. Yeah. So, can I get him back to the models and/or the app side of it? So, you know, [clears throat] how does you you were leading data science for a few months, and now you're over all completions. [clears throat] What does the life cycle look like? How are you guys coming up with ideas for different models to try to work on, or to test, or, you know, apps to build? And then like, kind of what does that look like? All right, so you've got a concept. Do you have a team of data scientists that kind of work on it? And what tools are they using there? And then how do you guys kind of roll that forward to production? Yeah. Um we do have an R&D team, and within the R&D team, we have a data science team. So, it's a combination of subject matter expertise in the oil and gas industry and data science expertise and machine learning expertise. So, I think it's it's kind of like the best combination you could have because that is one of the challenges of doing data science in this industry, that you can't just look at the data on its own without understanding all of these physical behaviors that are happening underground. So, um we do have either a research or data scientist or machine learning engineers. We have uh sometimes prototyping, sometimes, you know, they're parts of different phases of it. Um so, um the way it works is usually uh prototyping for a new model, for example, if we already have like a baseline model, for example, for like predictive drilling, and it's more like we want to expand to a new base, uh or a new use case, um then we already have all of these data in Corva collected. We have like our own way to build our feature. Uh we have a feature source, so you have a way to build your data set, uh train your model, and um test it. Maybe have other teammates uh confirm, validate uh the results and then we also check it with our other internal SMEs. Okay. Um, just to make sure that the results make sense with what is happening. Um, just basically the values that it's predicting make sense in real life. Yeah. And are those, you know, is that work being done primarily in Python for the most part? Python, yeah. And all of our deployment lives in AWS. Okay. So, yeah, and all of our models are Python. Nice. So, I guess talking about AWS, I mean, do you know, I mean, are you guys using like Sagemaker or any of those things to help deploy stuff or is it more like homegrown like containers or all of it or All of it. Yes. A combination. Yeah, so we use Sagemaker for some of the training and some of the deployment. Um, our feature storage is in S3. Um Yeah, pretty much. Nice. Yeah. So, and then, you know, we once we have a model that offline looks, you know, in theory looks good, then we deploy it and run it in shadow mode, meaning like dark launching it. Basically making sure that as it's predicting it's not necessarily controlling anything. You can see the predictions, but then you can compare how that compares to real life data. And maybe test almost, right? Yeah. Um, and once we have confirmed it works, you know, that also helps like making sure that the data connection everything works. Right. Um, then we launch it for real in our production environment. Nice. I want to ask you a question that applies both to your current position, but also I know just to your a lot of your past roles, but what are some of the biggest kind of hurdles and wins that you've kind of seen or best practices with, you know, real-time, quote [clears throat] unquote, high-frequency kind of completions data, right? Whether that's, you know, from your time at Well Day Labs or Corva [snorts] um, you've dealt, you know, you're in a unique position that you've dealt with a lot of this really ugly junky unstructured unstandardized data sets. [snorts] And you know, all the people that don't mess with that or even just like you know, have different they might have an analytics background, but they're dealing with production data or drilling data even, right? They're not they don't have that ex- direct experience seeing like how junky and ugly and unstandardized this data can be Yes. and how off like how fast it comes in relatively speaking to you know, a SCADA type data where it's 15 minutes or an hour or whatever it may be. [snorts] Oh, I should Where do I even start? Um [laughter] Um [snorts] Do you have any like protocols or databases or anything like that that you've used in the past that either worked really well or did not work at all for some of those things? Um uh not on top of my head right now, but um challenges like so many missing channels, uh missing data, changing channels, um not synced data. So, if you're pulling from different vendors, they all have different clock times. [clears throat] Um that can be a very challenging one if you don't have a real record of truth to compare. Like uh especially for example with offset pressures, that's a huge challenge of like and the data comes in different frequencies, right? Every 15 seconds, every minute sometimes. And even if they're one second, they're not synced. And that was a big thing that we were trying to solve at Reservoir Data Systems was that we were able to time sync everything Yeah. together or from the same clock. I think there are a lot of different techniques to fix a lot of those. I think the biggest challenge that I see sometimes is when you're actually missing completely data points. Like your FR is not coming and then you see for example, you're trying to analyze the data, you see pressure changes and then you don't know where this is coming from. You think it's happening down hole, but it's actually you're just not seeing that data. FR two ran out and now they're not pumping any FR and the pressure's walking up and they're [clears throat and cough] Yeah, that's kind of an extreme example, but I think it does happen a lot of times where you're missing chemicals or the acid or sand or whatever. You can't recreate it necessarily. Right. And then you're kind of blind to figure out what is what is happening. So, I think that again, if we go back to the the initial conversation on completion has so many variables even missing one or two, it can completely misguide you on what's going on. Yeah. So, that's I think we ended up just basically saying read all of the channels even if they're not mapped necessarily in the van so that you at least have Yeah. Yeah. So, when they do switch FRT in the middle and they hadn't mapped that channel over to stream to you, at least you captured it. [laughter] Yeah. And I think something that we have a Corva for example that is great is like we have a lot of ways to confirm that. We have our up steam 24/7 and they're always confirming like, okay, is this for like are they really not pumping this or you know, we can check the design, the data coming in, the post job reports. All different source of our data and then I feel like I know Corva was going down this road at one point, but do you you all have your like say your own like hardware that's reading and we were trying to do it kind of light and do it as light as we could at RDS cuz we were just a small team, but I mean whether it's you guys or or even like I think Coldbore and like you've got people that are looking at the data 24/7 and then it's seemed like it's almost a necessary component to do it really well. Um otherwise it can get away from you really fast. that's the operator wants good like they will tell you and I know this because they told Bobby and I that when we were doing it. I want I'm paying you for good clean data and then it's like, okay, well, I'm just the data broker. The guys in the field are the ones that are actually control the data that I see to a certain extent. So, it's like at some point you have to potentially get them involved to make sure that the guys in the field understand that Yeah. This Yeah. This has to be there or I mean it always comes back to the whole people, processes, and technology thing. Like I mean you can't just say like, oh, we're just going to capture the data and it's it's good and all you if we we got to have some processes and ways that things are done or else you can't Yeah, but then you also need the right technology and one thing I wanted to get there is cuz I feel like I know Corva was going down this road at one point, but do you you all have your like say your own like hardware that's reading the frac van? Or cuz I know I think at one point they were getting some data from RDS. I mean that sound like they'll get it from wherever they can get it, but um I think they were working on their own hardware as well. Yeah, so we do have our stream boxes for completions. But we also are working on our WITSML feed intake. Yes, so we started to um we're already rolling it out actually. But yeah, so the idea is to be more flexible, have more flexibility to be able to consume more and more data sources. But in the cases that there aren't any, then we bring our stream box. Got you. Yeah, I told people for the longest time that like that was the biggest kind of that was the trickiest part of the completions data like live streaming space, right? Is it's like you either have to have a box or you have to have a direct integration with the service company that's recording it. There's really no way around around it, right? And it's odd because even if you do have it with the service company, then you have a vendor that's pumping on their own or recording something on like you'll always have you need one sort of aggregation. Yeah. [snorts] No, you mean even just the timestamp thing that you mentioned, right? Like Bobby and I implemented UTC while we were at RDS because this became such a one you've got the daylight savings time, which is a whole [laughter] a whole thing. We won't even have to unpack that cuz everybody that's dealt with that understands that. But then two, it's just like hey if I can always say that it is pegged back to this time and all of our stuff is pegged back to the same server time, then you have a very high degree of confidence at least in in some source of truth. But that's the hardest part is like there's so many vendors on the completion side and they all have their own data sources and their own data sets and their own frequencies and their own you know naming conventions and stuff. And so that's uh yeah, I saw the hardware side for Corva as like a means to an end, not because you guys wanted to be in the edge hardware space, but it was just in order to provide the value that our software can we have to be able to get the data into it, right? Exactly. And you have some control over it at the very least. You have some control. I mean, like Yeah. Do you see Are y'all starting to see [clears throat] more and more people on the service side being like, "Hey, you know, we can give you an API and MQTT with some L kind of feed so that we don't have to have a box or people still very kind of closed best with that?" Yeah. I don't think I have an answer for that. fair. No, I mean, it's I've always thought, you know, the service companies have like have the best opportunity to excel in this space because they literally control the most of the data at least, but none of them have So, I think it goes it can go either way. I think we have some partner service companies or vendors that are really good at working with us to get more and more integrated in a long way to get these data and some others that don't. So, we still have to stay flexible for all use cases. So, now I mean, you're the completions product is it is it Does pretty much everything we're like living in the cloud or do you have like kind of edge modules or people like say in a disconnected environment are you able to still to It's all in the cloud, uh but it is in our plan to bring it on prem. Nice. That's awesome. That's Yeah, no, it's It's thing it's coming. Yeah. I was way too early at Hivecell, but it's coming. Um It's I mean, even just the stuff that you're talking about with the drilling on the drilling side, right? Like being The fact that y'all have the ability to have access to control systems to another operator or service provider is huge in my opinion cuz most of them are very like "Nope, this is mine." Yeah. But I think the industry eventually has to go that way, right? There is no other way. They're like everyone is an expert something and that's how we move forward collaborating. So, kind of you know, maybe tying some previous experience out like well, I think we'll skip a second and we'll come back to you kind of being out of the industry for a little bit, but um can you kind of compare and contrast you know, how you all were solving problems at Well Data Labs versus what the way Corva is doing it. I mean, and some of it may just be the fact that you guys were doing it 5 10 years earlier than them too and that was the best best way to do it at that time. But kind of compare that and then just how that experience has helped you, you know, in this role and Yes, I need to think about it. No, there's there's a lot to unpack here. It's [laughter] a big It's a pretty broad question, but Yeah, I think just like [clears throat] anything, you know, it was several years ago. So the industry was a little different. Well, they started more as a post job analysis and then moving to real-time core where I started with real-time. So I think it's just also kind of how their mission and vision worked a little differently. Yeah. The industry is way more open now and understanding of the importance of these data. Operations are getting more and more complex and you really need to look at these data in real-time. But I think that a lot of the work that I did there kind of opened the door to for a lot of people, at least in completions, to even look at these data. Like that the example of the water hammer paper. That was I think one of the first in a long time talking about how can you use this one second data because it's not something that people were using in completions. Right. It was kind of living in an image somewhere in a PDF and nobody was really looking at it unless you're like modeling, which is rare these days. At least in unconventional. Yeah. So, um, yeah, I think it ties really well, but in in I guess in different I mean, I guess in both cases you were still using the high frequency data, but it was just more coming in a batch in a way. You know, from those but the post stage files essentially were getting loaded and mapped. Um, the thing is too, I mean, there's probably still like there's still a lot of companies out there that I think we would be one of them like where it's we don't necessarily I real-time that necessarily. I mean, like but it'd be post stage but like you want the data come with the high even just using the high frequency data thing is still something that is not a lot of people probably doing enough of. Right. Um Yeah, that's the the craziest thing about the oil field to me is a lot of times we capture all this data and then it just goes to die somewhere. We're just like like post job reports, right? Like forever as a field engineer, I was generating post job reports and it's like, oh here's the ISIP and all these important things that go into a PDF and then Yeah. I think so I agree but I also understand why people don't use all of that data like we're not data analysts, we're engineers. [laughter] Sure. And I think that we're just collecting all these massive data sets and I think a lot of engineers had a learning on the way how to even deal with it. Um and I think we've gotten a lot better as an industry to look at like critical views of this data, how to compare it, what does it mean, uh what are all the variables involved? Um What is, you know, statistically something that matters or not? Um and I think that I go a little easier on on on that end on the industry because I I don't think we are necessarily we don't graduate like looking at dashboards and how to build dashboards. Like it's not something like we understand the physical behaviors and then we go to the field and learn about operations. You're not necessarily learning to analyze data. So, I mean, it's kind of part of the skill set of an engineer to have, you know, critical thinking and problem solving. So, it goes in into our skill sets but it's not necessarily something that we out of the the box coming out of school like are necessarily good at it. So, I think it is a good thing that we are collecting all this data and eventually we'll find ways to make it easier for Yeah. I think that's what Well Data Labs and Corva and a lot of these you guys have done, right? Is it's like as an engineer, I knew there were so many areas for optimization on the frac side and in completions, but then at the same time it's like there's so much data. A human can't look at that data on that type of scale and draw lots of great conclusions. in operations, then you don't have the time to even sit there and really like dig into it. I mean, you probably intuitively you know certain things just cuz you've seen it, you know, stages hundreds of times, but um But yeah, having those tools to allow the average engineer to make sense of it, I think it's such a strong and so important way. We talked about that a lot recently, but like can like, you know, if if there's an algorithm or something that's doing it, it's doing it consistently, you know, and you may not be perfect, but it's going to be done similarly where you have less bias and interpretation, whereas like, oh, you know, cuz [snorts] I I I think one thing and we can get into this is like you know, you've got the night crew versus the day crew. And like even just comparing like, well, you know, Jim is a lot more aggressive than Bob, you know, but you know, so he's getting to break down way faster, but he's over capitalizing on horsepower, but you know, other one are taking way too long Oh, for sure. I used to have consultants, you know, we were getting stage bonuses back then, right? So the treatment in the field well was 100 barrels a minute and they would run up to 120 barrels a minute just so they could get more stages in in the day, even though that's crushing all the equipment, it's not what, you know, all this fun stuff, right? And it's like, [laughter] yeah, there's there's consequences to doing those things. Yes, you get more stages in per day, but does your equipment break down faster and then you're shut down doing maintenance for more often and all of that fun stuff, right? It's [snorts] uh yeah, it like it there's just so many variables in all of that stuff. But I mean, I think one thing and and again, I I I'm still kind of bullish on just like post stage cuz most people don't aren't watching it and making, you know, even still like we were trying to preach maybe you can make diverter changes or back down completions while you don't know what's I don't think anyone's still really doing that for the most part. Like doing interstage are, but yeah. On a whole, I don't think I mean, I'm sure hopefully there are some people that are pushing that envelope or trying to do things more intelligently where like, oh, we're seeing some response on this offset well, let's maybe back it down or do certain things." Um but I think one thing that we underestimated, and I think a lot of people do, you know, I think probably even Well Data Labs going from being in a the batch thing where you're getting these nice files that start and end at the right time is like just identifying certain parts of a stage like from a streaming data set or what well am I on and what time did this actually start? And then like All right, this shut down is Is this a new stage or is or did they just shut down? Like all these things that are not into, you know, don't actually come across in the streaming data. Yeah. Um unless you set some processes and stuff in place for that. Yeah, we do that in real time at Corva, and basically recognizing where pressure testing, where pumping the pad, where pumping proppant, we finish, recognize size and be all these different activities. Is that done algorithmically or are there uh people like in ops that are helping with that? It's algorithmically, yeah. Yeah. Okay, in case there are issues, and you know, every now and then the algorithm will get a run if the stage cuts early or data cuts early or whatnot. So, uh we would um kind of QC it. Nice. Yeah. No, that's uh I mean, Bobby, you wrote you wrote a little script to or model to start doing that, but it's one of those things that it's uh once you have it, it's so easy to at least theoretically to back out all those aggregation, you know, you could generate your entire post job report basically once you Yeah. all that stuff without a And that's that's what feeds all of our dashboards, right? Like all of these activity detection algorithms feed into our dashboards so you can see when you, you know, you can automatically start seeing swap times, uh every time like if you are comparing like trying to optimize uh time in the pad, then all of those visualizations get um pulled in right away. Yeah, that's awesome. And I I think you mentioned earlier, are y'all doing like you do the frac, but also like the water line side as well? So then you're down. Okay, and pump down, okay. Mhm. You got the whole thing. Well, and that's that's how it should be done, right? You have to understand all pieces of an operation to really understand what the data is telling you. Otherwise, you don't know. Yeah. Yeah. Do y'all get into drill outs at all yet or anything? We have a solution for it. Um it's kind of in early stages. Okay. No, it's [snorts] I think that's a natural progression, right? It just keeps moving down the line until you get into production with SCADA and then you've got a full data Well, I mean like I mean our little thing our RDS we're going to try to go capture everything from first perf to first production. That was like kind of our our idea. Like so. Yeah, you can steal that. No one else [laughter] wants to use it. Uh trademark trademark. Um [clears throat] but okay, so let's look All right, let's talk about like I mean like well data labs but like kind of what led you into well data labs or coming Let's I mean now is a good time to start You know, what is your background cuz I were talking I think we're hinting at some of these stuff. But what is your background and where you from but how did you get into tech and Yeah, absolutely. So I mean I can go way back. So I started as a frac engineer. So I'm originally from Venezuela. I did my undergrad there, worked in the field there in Ecuador as well. Mainly offshore in Venezuela. I was a stimulation engineer. So I did like fracs, acid fracs, gravel packs. All sorts of anything that you can pump basically. Very different world though offshore than onshore. And then conventional really. You have three different worlds. [laughter] You break into zones. Yeah. Whole another can of worms. And then I worked more in like I guess in the design not in the field like more in like in the office I guess project managing and that was I moved to Italy. I worked there with Baker Hughes. Yeah, it was in completions but more on the completions tool side. That was really cool as well. So frac and frac and packs. Okay, nice. Like different places in South Europe. Um and then I quit my job to move to Colorado to attend grad school. Nice. At Mines or Yes. So, when I was at Mines, I was collecting a lot of data for my experiments and I kind of had to figure out how to deal with it um and how to even make sense of it. So, that's when I started kind of more into data analytics, uh learning some tools like Spotfire, uh you know, we're using MATLAB, some basic things. Um at that time when I was uh at Mines, I started working at Well Data Labs uh and I was doing sort of a lot of different things. Um a lot of research, uh understanding the data, um partnering with customers to understand their data, publish papers. Eventually, I launched the data science team there. Uh we built a lot of uh the data science cup really cool stuff really early. Thanks. Uh yeah, we did a lot of a lot of uh machine learning and uh data science applications. Um yeah, I was there for over 5 years. So, and so it was you know, I can we like to talk about the the tools and tech there. So, I mean, was it was it Python even then or was it any R or anything or No, we didn't really I mean, R is great for statistical analysis, but when you want to deploy things in in a product like environment in the cloud, then Python really is your your best solution right now. Um and even back then. So, yeah, Python, SQL, AWS pretty much. Nice. Okay, so had experience with AWS um both stops or we'll talk about some pri- some of the stops in between as well, so. Yeah, in between as well. So, uh yeah, and during that time I was also very involved with SPE, which um a lot of different things in the technical side and conferences, speaking engagements, and whatnot. Um eventually, I felt like I was plateauing in a way, Mhm. which is odd because I was a distinguished lecturer and um I was publishing and I was doing a lot of cool things. Uh, but technically, I felt like in the technical side, I felt like the industry wasn't there yet, and I wanted to uh explore or learn more about like the data science and like the software uh pure software um shop uh kind of view. So, that's when I went to Vroom uh outside of the industry. Um because what Vroom is. It's an e-commerce for used cars. It's similar to like what Carvana is. Yes. Yes. So, very different. like it's all online, right? Yes. So, you buy your car, it gets shipped to your house. I think one of the guys in my neighborhood used Vroom. Yeah. But now they pop up on whenever I He had a hard time getting his title from Maine, though. Took him like a year. Yes. That was a huge challenge. [laughter] I can't imagine having to ship like like I I I gave him crap for I'd drive by every day for like a year, and it's like [laughter] he still had like a Maine license plate. I'm like, "Dude, you live in Texas." Checks out. Um yeah, that was a big challenge. I mean, it it was a very interesting and complex uh like business. Uh A lot of people Because it was that on the buying and selling side, Yes. So, like almost like what Zillow like what kind of got in trouble like you know, you start buying at the wrong prices and cuz I know Vroom and Carvana have all of them they they really took off during COVID and then had some, you know, In buying and selling, it wasn't necessarily a challenge, I think. Just that the titles and registration was a big challenge. Logistics was a big challenge. I mean, shipping cars from New York to California. Yeah. And like car shipping is a whole can of worms that most people don't really think about in town that are there was a lot of data involved. So, um yeah, the the technology team there was about 300 people when I joined, and like the data team, we were like 25 people. So, it was really cool. It was a completely different experience, more like bigger tech, if you will, where you have an entire team for like data analytics, one for data science, one for machine learning ops, one for dev ops. So, it gave me a lot more perspective on that and more experience. So, I really enjoyed it. It was very out of my comfort zone. That's where growth happens though, right? That's the the best part. And I'm like that. I like to push myself out of my comfort zone. I know that's how I learn. And I knew that's what I needed to do to get my mindset out of the subject matter expert petroleum engineer to a tech, you know, Yeah. data science tech. Yeah. I would never be Oh, I don't think pure data scientist because, you know, I'm in between the two and I love that. a balance. Uh But, yeah, that was kind of my my way to to learn that. So, um What are some lessons that you saw that they did like really well that you think the energy or oil and gas industry should can borrow or or take from it? Even Or that you even brought with you. Yeah, as simple as like structuring the teams and stuff, right? Yeah, structuring the teams. Uh it really gave me a perspective on like a lot of different roles I've never heard of [laughter] at the time. I mean, also I think data engineering and analytics engineering was starting to as well. That was That was only like 3 years ago, but it was starting to get more and more popular. So, um really interesting to see where one breaks and the other starts, if you will. Um I think it gave me a lot of more perspective on deploying models in real time and integrating with different tools. Like some of our models will take data from our website, from Salesforce in real time, make predictions, and the user you know, probability to convert, for example, Okay. and feed that back to Salesforce or other tools. Um Yeah, web analytics are wild. Yeah. People don't know. People The average person has no idea how much [ __ ] that is tracking them and watching their session That's what you talk about the velocity of data. I mean, like you know, 1 second data, you know, stacks up pretty big and oh, I can't use it in Excel, but now you now you get web analytics, click stream stuff, and it's like It's a deluge of of data. Even just the basic direct GA data is so there's so much web analytics data just out of the box. Right? Understanding, yeah, user flow of where they go, where they drop that that can get very complex and we had like 2 million visitors a month. So yeah. Yeah. People that market like e-commerce, people they [clears throat] don't get enough credit like genuinely. There's so much data and analytics that goes into that and the companies that do it really well have such a distinct like advantage. Right? Like I even notice it like in my personal Almost scary almost as scary one. No, it it's true. Like I get you know, and oh, I conveniently get this email right when I'm thinking about buying this thing and it's like how does that happen? Right? Like How does he read my mind? Yeah, exactly. [laughter] Yeah, we were we were only using data that we get like clicks through the website. So it was way less invasive than some other companies out there. But yeah, I think what I learn a lot about is understanding some of these like big data problems in a different looking at it from a very different perspective, not quite like time series data is and then integrating different tools, learning more about just like the deployment and the ML ops and What are some tools that you use there or learned about there that you have either carried over or think would be useful for folks on the energy side of things? Like that's honestly really where this podcast came from. Bobby and I were talking about Fivetran. He told me about Fivetran one day and I was like, oh, cuz I was looking for like a marketing analytics tool to pull in all of our social channels, Google Analytics, all this stuff and just put into a database, right? Like I don't want to sit and have to write API calls every time they update their OAuth, which they do so often on all these social sites. And so he was like, oh, you should check out Fivetran and he's using it for, you know, their data and their databases internally or we were looking at it and then I'm using it for like marketing data, which I didn't even know like when he told me about it, I wasn't even sure if they would have it, right? But it's this like really functional cross- industry platform that you can use in databases and everything. Yeah. Things like that are just like floating out there that the tech industry uses every day that the average person on the energy side has no idea about, right? Yeah. Um what do we use that I really like was Monte Carlo is it's like a observability software. So, it tells you like that it's not coming in. Um yeah, basically you can create your own sort of dashboard for what you want to observe, basically. But, um very user-friendly. I really liked it. Um we also used um um DBT for ETL. you're using. Um also really nice, easy to use. Um we use actually for querying our database, we use Popsicle. Which was like a very young uh kind of user interface for SQL. Um it like What was the just of that? I mean, I feel like I may have heard of it, but I haven't You can basically run your queries and then have like little tables to visualize it. It was like a very nice interface for querying databases. an IDE Yeah. but you have some sort of visualization Yeah. Um What were y'all So, if you're using like DBT or using on top of like um Postgres or on top of Redshift or Yeah, Redshift. Okay. Mhm. Yeah. So, also AWS. Yeah. So, it was a lot of AWS under the hood. Yes. like Yeah. Yeah, yeah. So, yeah. And then after that I was out of work for a few months, and that's when I started doing the mentorship with my friend Carly at Rebel Data Science. Um and yeah, then I joined Corva. So, let's let's dive into Rebel Data Science. Like, you know, you know, what is it? I mean, I sounds like you and Carly came together and put that together, but like kind of what what was the what what's the mission, I guess? Yeah, she put it together. 100% credit to her. Uh she's incredible. We actually worked together at WellDataLabs. Uh and I was her manager. And then she went to work at Activision, uh and now she's a superstar. like a big-time manager. Yeah. And uh anyhow, she's been trying to get me to be more active on LinkedIn. She was just suggesting that I should, I have a story to tell, whatever. And I'm not really that into social media when it comes to getting so much attention and exposure. Uh but eventually I got there. I started posting and um joined her, and it's really interesting because we have some similar some overlap in our experience, but we focus in very different industries, very different things. Uh But really the idea is just to offer mentoring uh or mentorship to younger folks who are coming into the data world. Um in my mentorship sessions, I focus a lot on changing industries. Okay. Cuz um that's I've done that quite a bit or changing roles. Mhm. Uh kind of going untraditional paths. Yeah. Um or people that want to be a manager, you know, go from individual contributor to manager. Um a lot of times it's just like resume reviews. So, it's really just a little bit of mentoring to younger people who And I say younger people, that's just usually the people that seek our advice, but it's not necessarily Absolutely. I was a teacher and I went into oil and gas, right? So, Yeah. You were? Yeah, I was a high school teacher. Wow. So, I went into oil and gas. And He's the reason we have to do all those proofs. Yeah, Bobby. [laughter] But yeah, so I mean like yeah, just that making making those kind of I mean that's a extreme jump, but I mean like some people just and I've actually had last year or so had a couple people wanting to transition from teaching into data. You know, I'm sitting with them and like well, I mean maybe a data role isn't even the best first role for you. Maybe you need to get into an industry and if it's just a different industry, understand understand the problems and right. You know, you yeah, you that's great you did a Google certification in Python, but now like get into a the ground floor and like use that tool set to help people solve problems. And you know, when you start solving people's problems, then they like you and they keep you around or they promote you. So. Yeah, exactly. Yeah, so a lot of it is that, like a lot of people don't even know where to start, um how to even look where to look because there is a lot of buzz, too, about working in data and then there's Yes. [clears throat] But you look at the last year, a lot of that has dried up. I mean, I had a buddy that was a teacher and went to did a full stack web development boot camp and then like a all of a sudden, all these people from Google and everywhere else are getting laid off and it's like he's got to compete against them in the job market. Like, you know, he's landed on his feet, but it's like it wasn't quite the path that people carved, you know, were saying for the last 5-10 years. Like, "Oh, learn to code, you'll have a job, you know, maybe six figures." Like, not necessarily, you know, as much anymore. Which is what a lot of people or influencers or whatever are marketing. Like, "Oh, you can get into data, you can make a lot of money." Which is true, but there is a way to get there, I think. Yeah. And um yeah, a lot of people just like go, maybe take a class, they still feel lost. There are so many industries you can do data, there are so many roles within data, right? And then even within the roles, they mean different things for different companies, for different industries. Um so, I think it's just a lot of guiding them and mentoring them and like just offering advice. No, I think Bobby's right. I think people have been fed, I mean, you're both right. People have been fed this narrative that it's like, okay, here's the formula to go get a role at Google, Netflix, you know, FAANG, any of the big FAANG companies. And that doesn't really exist anymore. And even then, I don't know that there was like just through the guests that we've had on the podcast, I don't know that there is like a true formula that, you know, like if you do this, you are guaranteed to go get into this, you know, specific role at a specific graduate in top 10% of your computer science class at Carnegie Mellon and then you can go to Google or Apple and whatever else, but I mean, like You started at MIT when you dropped out. Like, maybe. [laughter] That's the I mean, I think there are easier ways to get into data. And there are like [clears throat] ways you can get there. It's just people don't know that all the time and they think okay, data scientist is a new thing. I should go get a boot camp and then find a job and it's not always that straightforward. Yeah. Um and I really like your suggestion of like always maybe find your industry first or maybe find something I'm always trying to make it gradual for people because well, I like to push myself and I'm like I go for like wild changes sometimes in my career. Uh not for everyone and sometimes if you're trying something new that can be very scary and just completely like yeah, scare you away from the whole thing. So um I think gradual things and and then finding overlaps in your strengths is always my biggest advice. It's like if you already have the skills like take advantage of that, market that and then use that to slowly get to the other side. Yeah, I mean I mean I was beyond uh lucky to to get you can get my foot in the door. I mean it just I was coaching baseball and one of the guys was like oh, you'd be good at this. You have a math degree and we we could we do well with math degrees and but then you get in and you know, it's like I to your point I had to lean in on all right, I was a teacher and a coach. What am I good at? I'm good at communicating, you know, I've got aptitude. I can learn things. I know how to you know, I'm going to lean into that but like I bought me and my whole elevator pitch is like I didn't know what a VLOOKUP was in Excel when I walked in the door and then they're like you're going to be the Spotfire guy but like just you have to be willing to like lean in and learn stuff but but honestly when I I wouldn't have been anything if I hadn't learned the petroleum engineering side of it, you know, and I was lucky to have be at Conoco where they had a lot of infrastructure and oil was 90 plus dollars a barrel. So they sent me to a couple trainings but you you've got to yeah, like you got to get learn the industry but then like you said lean into your strengths. Like what do you do well? Like what there was a reason they hired me. It wasn't because I was awesome at data when I walked in the door. Yeah. Yeah. Um but then I think then from there you carve your path. When I came in I think the idea was like oh, you know, maybe I'll work on a petroleum engineering degree in a couple of years. I can maybe get in as a production engineer or something like that. And then I got in and I was like oh, I kind of like this data thing. So like yeah, don't I think you can go in with blinders and say, I'm going to be a data person and you may realize you freaking hate it. Yeah. You know, like some people that's not for them. Like that's like the worst thing in the world to sit at a desk and like look at a screen and like try to make sense of data. It's not the sexiest job in the world, but Yeah. It's very irritating. You can pull your hair out quite a bit. I mean, I've lost most of mine now. So, it's it's yeah, that's I do agree with you though that like people think that it requires like an industry change or something, but it's like if you already have a subject matter expertise in whatever it is, how can you leverage that and maybe get some data certs or classes, whatever, so that you could turn around and be that differentiator where, you know, someone that is just coming in with a pure data role who has no idea about the SME of whatever industry that is or job that you're working at, you're going to have a leg up over them because you understand the context. You understand the problems. You understand what you're trying to optimize or solve for versus you know, a computer science major or developers generally, you know, they just go off like a checklist. They're just Yeah. Or some of the bad ones, but And I think the whole like sexiest job, whatever marketing, like really hurt to for people who will feel For example, I've known certain people that got into data and then they thought data scientist was the role to have and then they're like, well, actually like analytics better or like data engineering better, but it wasn't like I think the sexiest thing and then A, that I think there's just more jobs and more value in data engineering for most people. I mean, I think a lot of companies skip that step, right? They're like, we need eight data scientists. Well, the data you're feeding to it is dog [ __ ] Yeah. Yeah. I mean, uh Here's the 10 years of our frac data that's unstructured and has no naming conventions. I mean, that's one of the things Bobby, we met with a guy at at good size operator at and his his first language or machine learning model he was building was to go in and normalize the headers on all the frac data historic frac data because so that he could then go and build a predictive model using the frac data. And that's like, yeah, no one thinks about, you you know, you have to have good data and there is a huge, I think that's a big thing that was like, there's so many segments of data and analytics and data science that people just don't realize, right? Like they don't know about, yeah. Right. Like there's the engineering as marketed or they're not as, you know, I don't know, people don't think they're as cool so they don't want to get into it or they they feel, I don't know, like it's status-wise in your career doesn't give you the same status or what not. Um and I think that's hurting Yeah. these careers of these people that may have they might add a lot of value in these other branches. Yeah. You know, the biggest thing is be useful. Like that's just my like So make yourself useful, solve people's problems. Like it doesn't matter what, I mean lots of people you can call me whatever you want to call me. Like I'm like, I mean, you know, data science, data engineer, whatever. I mean like I just want to solve your problems and probably going to use technology to do it. Mhm. Yeah. But um yeah, personally I don't like to put in a box. Yeah, I mean like if I need to write up a machine learning model, I'll figure it out. Yeah, but most time probably SQL query is going to be good enough to get people the answer they need. Um you know, the most dangerous people I know are an accountant who knows SQL or or Python versus sitting there hammering away in VLOOKUPs. It's like Yeah. Well, they're going to do things a lot Yeah, more [laughter] efficiently than the other people like you know, same with I mean there's a ton of I mean engineers by nature are much more technical and they're curious with those kind of things. But when they have that context and then they can bring their ideas to life without needing to translate to someone else. It's Yeah. really powerful. I genuinely think it's just so much easier to train a subject matter expert on the data and analytics piece than it is to take a data analytics person and try and train them. It just takes a Depends on the industry you're in. It depends on the person. But like generally speaking [snorts] if you if you understand the context and the problems in the industry and what you're trying to do, then as long as you like to solve problems, you have a much Yeah. Yeah. But people don't think about it. They think, oh, I've got to have this cert or I've got to get a thing company for however many years and hate my life and then go do something else, right? I also think that the career of any data role that you have has been so much separated from the data that you're using and I think that's so important because that's something that I saw, you know, moving away from the industry. It was like, oh, there are cool and interesting problems to solve. It's very complex, for sure. But am I passionate about it? Not really. And that really drives like a lot of innovation. Like if you're not thinking about this problem constantly, um if you're not passionate about solving something, Yeah. like yeah, it's a it's a cool experience, it's a cool job, it's you're building cool things, but are you really passionate about it? And I think that's one item that I think sometimes people in the data world, careers, or whatever you want to call it, um don't always consider, right? I I think personally, because I'm so passionate about the industry, because I'm so passionate about completions, for example, that for me is very important and I've no I know it now. Um to not just work in data, but work in data that uh with problems that I want to solve and with data that I find interesting. impactful. Like that's what I love about where we're at in the energy space the last, you know, 5, 10 years and even moving forward is there's just so much opportunity where people like us can come in and make we've watched this problem happen, you know, the first half of our career and not get solved and it's like, hey, we can actually come in and make a change and change the way people do things and impact the company, whether that's making more revenue or cutting costs or whatever. Whereas like you go to a a really big, you know, major tech company and it's like you're optimizing some random algorithm that doesn't really change anybody's life or anything like that, right? [snorts] Optimizing ads. [laughter] Or the person that manages like the search button on the Google homepage. Like I changed the hex code by [laughter] a couple figures, yeah. Yeah, we didn't we didn't change the logo on this holiday, so. [laughter] No, there's it it's very true, but [snorts] are we man, are we already at an hour? Yeah. 56 Okay, we got a few more minutes and then we can jump into the speed round. Yeah. Oh, [snorts] there's the speed round. It's just a fun thing we do at the end to pick your brain. all the way back to I mean like I mean so how I mean being from Venezuela, I mean has oil and gas always been like has it been in your family? Has it been kind of your blood or were you like the first one to I was the first one, yeah. Uh my my parents are from Argentina. They moved to Venezuela in the '80s with my brother and sister and then I was born there. So like immigrant, you know, family. Um not really history on many things. Um so I was I grew up in Maracaibo, which is uh oil city. Um my state is it's a oil state basically. Um so it's just kind of was around everywhere. Um when I was in high school, we used to have or in middle school, maybe. Um the national oil company used to host this like um petroleum Olympics. Uh it was just like a reading comprehension kind of competition. Uh but it exposed me to the oil industry in a way that I could read about it Right. in a like sort of technical of what they do, what is it about, and I was found it super interesting. I would sign up every year for it. Um yeah, yeah, I should host it. Oil and gas Olympics. No, we I want to do a roughneck Olympics. Honestly, like get a bunch of drilling crews out and see how quickly you can rig up a frac crew and [laughter] see how I'm saying to put I'm saying to indoctrinate Doesn't sound safe. Oh, yeah. No, it's that's pretty true. Um and then frankly, you know, we had a lot of expats working in the industry and that was always, you know, having my parents from another country um I've always like find it exciting to travel and live in different places. So I think that was I mean I found it really exciting and interesting, but also the idea of traveling, living in different places, learning new languages, all of that was like probably the bigger drive on me getting in the industry. That's awesome. Now, I think that's a Again, it's one of those that like if you don't know, you have no idea, but once you learn, you're like, "Oh man, I could literally travel the world." Like especially early on when you're young, right? Like that's why I went to the field cuz I was like, "I want to go I don't have a wife, I don't have kids, I don't have anything holding me down. I want to go learn as much stuff as I possibly can." And the cool part about all these companies is they will send you all over the world if you want to to Yeah. learn all kinds of crazy stuff. [snorts] Were you always into kind of tech as well? Or are you just kind of like me where you love to solve problems whether and you'll find a way to solve it whether that means Yeah. You kind of found me. Tech found me. I'm like in your PhD was a big drive for that as well, masters, okay. It was I mean, I think I was just in the right time where at the right place, I guess. Um with Well, it all happened so kind of launching and me being in Colorado. I think it just kind of happened. But I've always been passionate about the industry and solving problems and just like both the technical and the operational side of the industry, so. Yeah. No, I think that's a a big piece of advice for those folks that are looking into get looking to get into the data science data engineering side of things is if you don't like solving problems or learning new things new ways to solve those problems, probably not for you. not the best cuz or dynamic work environment where everything is changing is fast. You have to think. Or yeah, then you the next week you've got to go turn around and redo the entire thing that you did the prior week because something else popped up or changed, right? Like it can be very infuriating at times, but it's super rewarding, too. Like being like, "Oh, yeah." Or even just like [laughter] I get this feeling every time I I pull an API in and I make sure I have the logging and you know, I throw it in GPT and tell it to put logging in and then you see all the rows coming in and you're like, "Oh man, I'm a hacker." And like you're not. You're just pulling an API. [laughter] Yeah, but even like you have some You know, people always point out I used to watch the shows and like, like, it doesn't [laughter] make sense. It's like, oh, this they're showing like a data scientist and it's like a bunch of HTML going down the page or whatever. It's like, Yeah, he's not he's not doing anything in DOS, right? [laughter] Uh, man, I was going to ask you Oh, what cuz you've worked at a number of startups and I feel like that's I personally love that startup role because it is so dynamic and there are so many problems that to solve and they do, generally speaking, make pretty impactful changes to the company or the user or whatever. What's some kind of, uh, you know, tips, advice, strategies that you've kind of learned along the way in the startup life to that, you know, people can kind of learn from? Now, that's a very big and open-ended question, but [laughter] [snorts] Cuz I mean, you were at Well, Daily Labs pretty early and you all you got to see the whole you got to see the whole you got to see the whole kind of journey there. Mhm. I think the biggest thing that you see in a startup is the different phases of growth, uh, and how the company will change, um, quite literally. Um, and I think for it's not for everybody. For some people, want stability or want, um, things to stay stay the same because they join especially when you join a small company, um, a lot of people have like this romanticized idea of like how things work when you're small and everybody gets along and everything is perfect and It's usually quite the opposite. People are like having disagreements, but you have you have private relationships to have that kind of candor. Right. Right. And then it grows and it's going to go in different directions and I think it's just being okay with riding that wave of, you know, You were you were six people. You probably didn't really have product market fit yet. You're probably still trying to find where you guys fit. So, it's like, you know, things just change like, oh, no, we don't even need this product line anymore. So, now that maybe this person or these things or we need to reallocate resources and Yeah, there's a lot of redirection, pivoting. So, being okay with change. [clears throat] Um, It's a really big one cuz I feel like I mean, I've been at, you know, pure startups from almost zero, right? Um you know, first couple of employees to coming into like existing companies that were more just kind of you know, call them family businesses that then want to go into more of a startup growth mode. And like people don't I don't think people talk enough about how during a startup's journey, like you have a very different you know, your first when you're under 10 people and you're first starting your prioritization is staying alive and being able to make payroll and, you know, find those first handful of customers. But then you get into like that growth phase where you're kind of or the crossing the chasm piece and stuff and it's like hey, we have some funding so we can go spend some money on some stuff. So, you have to literally rewire your brain of like, well, I could go use this cheap workaround that I used before, but we now have money to actually spend on these tools. So, should I just set it up the right way now? But you have to literally change how you do things. yeah. Yeah, and perspective changes. And I've learned to see it in a way that if I don't see thing change it things changing, I feel like we are complacent. So, I almost see it as a weird thing if there is no change. So. Yeah. Well, even like I think we're I think there was even maybe an article about it but when we were at RDS but like when you get into that mid-size company, that's like the almost the worst phase to be in, right? Where you're getting like the 20, 30 people cuz like now you're too big to just you know, move fast and break things like you did before but you're not big you're not big enough where you want to have like all these processes in place that are too bureaucratic but like to get to that next level you probably need to put some processes in place. Yeah. It's like that teenager phase. No, it's very true, right? Like I mean, every single company I've worked for that is in that has been a startup or trying to go into that growth, right? It's like when you're early stage startup, you just move and you're always moving and you're always changing, you're always pushing forward. But because you're doing that, you're not stopping and documenting or proceduralizing things or anything like that. And that aids in some areas but it hurts in others. And so, as you're trying to grow and you add more people and resources, it's like, okay, how do we proceduralize some of these things, but then that slows down how quickly you can move, right? And so, it's a very delicate like balance of that, for sure. Because you want standardizations so that you can scale, but you also don't want to crush the whole spirit of moving fast. That's I think that's really, really good advice right there, cuz a lot of people don't you know. They just don't know. Yeah. All right. So, we're going to speed around, yeah. So, we just pepper you with random questions about any mostly tech stuff and just give us a short answer just to learn a little. I'll try to pass. All right. Well, first one's not very techy, but it sounds like you're really well traveled. What's your favorite place to go on vacation? The beach. Just the Any beach? Okay. The Caribbean. The Caribbean, okay. Probably my favorite beaches, so right there with you. Um It sounds like AWS is your most experienced. Would that be kind of your go-to for cloud? Yeah. That is. [snorts] Um what about any kind of, you know, either social follows, whether it's, you know, Twitter, Insta, LinkedIn, etc., or kind of books that you've that are, you know, kind of tech related that were really impactful or Yeah. Um I think tech related follows, there is a few people on LinkedIn that I follow that I learn a lot from. Um And I'm going to butcher all of their names, but obviously Carly is, you know, uh Wasn't it Taylor? Taylor, yes. Carly Taylor. Uh Ben Rogojan. Yeah, I know who you're talking about. Um he posts a lot about data engineering. Joe Reese, he's also he wrote the book about fundamentals of data engineering. Um Books Oh my gosh. Um I don't know that [clears throat] it's tech related, but one of my favorite mean? Like even just personally like books ever is um Never Split the Difference. That's a great one. It is a great one. Yeah. Because I just think that in life and in work and in everything, it just helps you reframe how you think whether that is for negotiating or not, but everything in life is a negotiation. So, you're always negotiating with yourself, with other people, with your partner. So, I think it is just I love his approach to like open-ended questions and just I think all of that. Yeah. Yeah. I'm doing with my kids. Yeah. that I [laughter] was going to say you That was the the negotiations that you have with your children. You never You just don't know until you you have them and you're like, "Why are we negotiating about this?" But it's so true cuz he opened I mean, I vividly remember early on in that book, that's what he basically says, right? Is like everything in life is essentially a negotiation even internally, and I'd never thought about that before. And it's like, "Oh man, that's a good And then he says all That's a great book. So There's so much to be used out of that book. Yeah. Uh favorite open source library? Oh gosh. Can I pass that one? Sure. [laughter] I'll do one more. Um What's some advice This works perfectly. What's some advice you'd give to someone looking to get in or coming into the energy kind of tech space that you wish you knew Yeah. originally? Um Maybe generic, but just being curious and staying open-minded to new ideas and new tools and new approaches. I think it's just the way to get into something new. Just always being open-minded and curious. Like not approaching something from I know how to do this already. This is my way. Uh but having curiosity to explore other ways, and maybe you'll come back to your way. But always staying curious into what else is out there. What can could doing? I mean, I love when people bring in say like people from other industries and like other you know cuz I mean I was you know Liz and then with Biota or stuff where they're using like, you know, Yeah. those kind of concepts, you know, to model, you know, Yeah, there's just there's different ways of looking at things and if you bring people diverse, you know, backgrounds like they might look at it a totally different way and crack something that you couldn't cuz you had your, you know, your blinders on. So, no, it's very true. You get so Oh, I should have mentioned Liz. I love her. Oh, yeah. She's amazing. great. Yeah. No, that's uh Yeah. It's very like that's I'm still very much so of the opinion that generally businesses all have the same high-level problems. The industries are different. The constraints that you have based on the industry are different and all that stuff, but businesses have generally the same problems and so just being able to figure out where you fit and solving those problems can really kind of stand up help you stand apart. But, yeah, it's just that curia- you have to have the curiosity to go jump into it and not just be scared of like, oh, what's this going to what's going to happen from this or whatever. Or just being so stubborn or stuck to your way that cuz a lot of times being open-minded mean being exposed to other ideas even if you think or approaches that you've never tried, that you never thought of or that you think they're not going to work just being open to, yeah. Yeah. What Well, and then it builds buy-in from the team that you're working on, right? When everybody's voice is heard and their opinions are heard and then, hey, if they come back around to your idea, great. If not, not a big deal. Yeah. [snorts] No, I I think that's very true and that a lot of people in our industry need to There's quite a few that could learn from that, I think. Yeah, no doubt. Well, awesome. Uh where can people find you? Something like LinkedIn? Yes, LinkedIn. Uh just Jessica Uriarte. Mhm. Awesome. And usually there. Thanks for coming in. We appreciate it. Thank you so much for having me. Yeah. Glad we got you in here, so. Yeah. This is so fun. Cuz you're in from Colorado. May- People may have picked up on that, but she does- doesn't live in Houston. Yeah. No. She uh is graceful enough to fortunately Corvus is across the street, so we were really able to steal her for [laughter] a for an hour while while she was in town, but you did get to enjoy all the uh wonderful weather that we got last night. I I hope that didn't wake you up like it did us, but no. Oh, yeah, there was a crazy storm at like 2:00 a.m. Yeah, I woke up to my son inches his black shadow [laughter] face inches and then the lightning from the the windows behind him lighting him up. It scared me awake, so It must have been a well-insulated hotel or Yeah. [laughter] Well, I think it was further west and it just kind of moved up and over the west, but yeah. Nonetheless. Anyway, thanks so much for coming. We appreciate it. Thank you. Thank you for having me.