AI for Smarter Land Records & Data Quality

Full Transcript

ELIZA WITH PANDELL Welcome to this Pandell Leadership Series webinar. My name is Eliza. I'm the Manager of Customer Engagement and Training Services here with Pandell ESG. So, today we are lucky to have three speakers with us today, and I'm going to introduce them all to you now, starting with Michelle.

Michelle is the Manager of Land Records Services and has been with Pandell ESG for over nine years. She has 20 years experience in land data management and is considered a digitization expert within the energy sector.

Then we also have Dean. Dean is a Senior Account Executive at Pandell ESG with over 30 years of multifaceted energy experience. During his career, he's consulted for over 100 oil and gas producers.

And we are very lucky to also have guest speaker today, Trudy. Trudy is the Chief Executive Officer of PPDM Association, the energy data professionals, a global not for profit society focused on building a global community of data professionals. Trudy has over four decades of industry experience and is known around the world for her outspoken advocacy of data standards.

So, really great group of people here today, very knowledgeable. The topic should be really interesting. So, we are going to dive right in, and you can take it from here.

DEAN WITH PANDELL Thank you, everyone, for attending today. As Eliza said, we really appreciate your involvement here. We're going to talk about a couple of things that are really close to me and that's land and data.

And so one of the things that Ernst & Young came out with a study, and they have looked at two things, two areas that AI is really going to improve, whether it be energy producers, whether they be oil and gas, they might be solar, wind, but they're looking at predictive maintenance for heavy equipment and assets, and as well as intelligent optimization of operations performance. And I can attest to that too. In my previous life, I did a lot of work with the field operations groups. And it's an area that a lot of money goes into, obviously, for say, an oil and gas company. And it's very much needed to improve that area using tools like AI and digital, digitalizing your records.

So, some of the industry challenges that we see a lot, we work a lot with land records, obviously, and we see companies out there that are still utilizing paper land records. You know, the challenges with that is data is in a non-usable format, it's on paper. And we also see a huge resistance to investing. And more so because investing in trying to get out of paper is a big cost.

And you know, you look at the number of agreements a company has, whether it be agreements with landowners, whether it be agreements with partners, whether it be your well data files. The challenge is there's a lot of agreements. It's not uncommon for a company to have multiple data rooms. And so, it's a real challenge. It could be upwards depending on the company size; it could be well over a million dollars to get out of paper.

This is an example of, you know, one of the file rooms in a company. And again, you know, these companies may have multiple data rooms. You look at this and you say, well, what do you see here? Well, you see someone going in trying to pull these files. And these could be contract files. They could be joint interest files. They could be surface files, mineral files, or well files, right? So, you're going in, you're pulling these files, you're taking that time to do it.

And then you're taking the file back to your desk, you're reading through it, and you're trying to pull out the pertinent, pertinent information out of there. I also see a lot of floor space, you know, companies are paying for space, you know, what companies I used to work at, they used to have to reinforce their file room floors, right? So, they'd pay money to reinforce the floors so that they could have their paper files. Unfortunately, the cost is, is kind of prohibitive to get out of this. And so, we're still seeing this a lot in the industry today.

I'm going to introduce Michelle over to you. And this is one of the projects that Michelle's team is working on. It's for South Bow Corporation. If you don't know who South Bow is, they were a split out of TC Energy. And so, the oil assets went over to South Bow.

So, one of the main oil assets there is their Keystone pipeline. And so, that spans across three provinces here in Canada, as well as eight US states. And Michelle's team is really dedicated working on this project for them right now. And it includes over 5000 land agreements. So, you can imagine all the different pipeline segments and facilities that are attached to those pipelines.

She's been working with over 30,000 PDF agreements and trying to digitize those for South Bow. So, Michelle, a very good friend of mine, she is from probably my favourite name for any town in the world. Originally from Sugar Land, Texas, and now is based out of Houston. So, I'd like to introduce Michelle Uzick to you.

MICHELLE WITH PANDELL Hey, everyone. As Dean said, I’m Michelle Uzick. I actually have experience going through those file rooms. So, I just wanted to add, I've been there. I've gotten many a paper cut over the years. I worked for an oil and gas producer actually out in Sugar Land, on and off about 10 years, and they still haven't digitized their data. So, it looks like that file room, but a little worse.

So, we'll talk about data. So how do you get access to all of this important data? By converting your unstructured data to structured data. Instead of using methods that traditionally takes days or weeks of manual effort can be transformed through AI powered document processing. But not all AI solutions are created equal, especially when facing unexpected document formats or complex requirements that go way beyond basic extraction.

Now that we understand the challenges, let's explore how intelligent document solutions can transform this reality. Starting with foundational capabilities and then moving to the advanced approaches that handle those unexpected situations.

Let's start with the basics that come after the scanning of the paper land right files. These four AI capabilities listed here have already revolutionized document processing and energy sector. While straightforward, they deliver substantial efficiency gains by automating previously manual tasks, forming the foundation that more advanced solutions build upon. The first phase in the process is to OCR the scanned documents, which is converting image-based files into searchable text.

So, next is the separation of PDFs, which consists of dividing land agreements into their components, such as the primary land right document and any supporting documents, such as amendments, chain of title and so on. Because oftentimes when you're scanning, you'll scan one big PDF containing everything related to that land agreement and you want to make sure that you're splitting them out into their individual components.

Next up is a de-duplication tool that we have, which is removing a redundant copy to prevent processing the same document multiple times. As many of you probably know, there are going to be file duplicate files in different field offices. You know, you scanned and faxed something to someone at one point 20 years ago, maybe 30, 40 years ago, and you have all these multiple copies that you don't need. So, what we do is we use our tools to identify those things and combine, confirm and combine those into one PDF. We will, you know, nobody throws away your PDFs and that's at your scan. You still have all the copies that are just put together into one.

Classifying of the documents, which is to identify types and apply the appropriate extraction rules.

So, the final step in this process prior to the importation of the data into a land management system is extracting data, which is pulling structured data or information from the documents into a usable format. In the next three slides, it's a little bit of an animation. I don't know how well it's going to show through here, but, Dean, if you want to press through those, see some pretty colors.

[Showing various data within documents]

And then we take all of that data and then we import it into a land system for you.

So, what if AI could do more than just the basics? Here are six challenges that go beyond standard document processing. Each one represents a potential project stopper that an adaptive AI approach has turned into a manageable solution.

Here, I'm going to speak to the first three of these in the interest of time. So, in the worst cases, we have things such as missing first pages, documents that lack critical information, entity recognition, complex party names, meaning separation and standardization, and inconsistent formats, regional document variations requiring adaptive classifications.

As you all, again, probably know you have documents in many different provinces, states, counties, and they all have different formats. So, we'll take a data set and be able to process them knowing all those differentials in those types of documents.

And here's an example of some of the code we have used to perform these tasks.

So, now that we've explored the complex challenges our adaptive AI solutions can overcome, let's examine the three key roles AI plays in transforming the document digitization process from problematic to powerful.

Finding nonconformance. It identifies discrepancies, missing information, and unusual provisions in land right documents that require human review. Handling routine tasks. The AI processes thousands of documents for classification, extraction, and cross-referencing in hours instead of months freeing your land team for other important projects. And augmenting the human expertise. AI handles routine verification of flags potential issues allowing land professionals to apply their judgment and expertise to complex decisions rather than data gathering. Together humans in AI create a powerful synergy that maximizes both accuracy and efficiency.

DEAN One of the big things around AI that we're seeing, there's so much data coming out now that data governance is significant. So, everyone knows about the cybersecurity challenges across the world. You know, the strict data isolation is becoming more and more important as these data models get built out, the AI models get built out, AI agents are using those models. We just feel that it's more and more important to make sure that you're using a certified vendor.

So, you know, some of those certifications include like ISO 27001, which deals with how you deal with data. And then also SOC 1 certification is you have to get an independent audit every year to become SOC 1 certified. So, just really important to...that I can't express the importance of that enough.

So, my next guest here that I'd like to introduce, I've been familiar with PPDM for, oh, I would say over 30 years. That's when I first got familiar with it. And I thought it was a really, really great cause because... and great initiative. Because at the time, 30 years ago, we had access to data. It was somewhat limited though, in comparison to now. And we really were struggling with trying to pull multiple data sets together and trying to, you know, similar problems back then, but we've solved a lot of those problems now with the tools.

So, when you look at PPDM, back in the day, it was called Public Petroleum Data Model. And the initiative was structuring a data model across the industry, you know, specifically around oil and gas at the time. So, they would work with wells, facilities, pipelines. But there are three pillars where data community, data resources, and data professionals. And we're very lucky today to have Trudy Curtis with us today. Trudy was, she's current CEO of PPDM and one of the co-founders back in the day. And so, I, you know, like I say, a passion of mine is data and a passion is land. And I'm very proud to introduce Trudy to you.

TRUDY Thanks, Dean. I really appreciate that. It was a great talk. And the representative of the kind of things that a lot of industries trying to do right now, we're really looking for ways that we can implement and embed AI in ways that are going to help our business processes to be more accurate, more timely to help us to make more decisions, more accurately, etc.

One thing I'm finding is I really want to emphasize now is that AI is actually not magic. And you might have some managers in your company who think that AI is magic. And it's not. It's like a really intelligent child who's got a severe case of ADHD, who forgot to take the pills that day. And there's lots of things that can trip it up. And those are the kinds of things that PPDM members work on. So, we're not for profit professional society. And we work to understand where problems can arise with data and to figure out why they happen and what we might be able to do with about those problems working collectively.

So, as Dean said, one of the things we did many years ago was to generate a very large, very complex relational data model that talked about over 63 different subject areas in the business. And what we found now is that people want to use different technology. There's so many technology opportunities to open to us now. And we can move between them very easily and flexibly.

So, now we think about what is going to happen as we move into this AI space. And the paper documentation that we were just looking at now has all of the same interesting issues and challenges that we do with people. Documents can be very individualistic. People write what they write. They use the language that they're used to. Even things as simple as date formats are really variable. And training an AI to understand and recognize how those are working in each individual document is interesting.

I've worked with paper documents where they transition date formats from year, month, day to month, day, year in the same document. And that can be a very difficult thing for AIs to handle. Units of measure can be really tricky for people. So, I don't know how many of you have seen some length measurements that are done in Faraday's. Obviously not a length measurement. But these kinds of things we work on at PPDM.

So, when we think about the kinds of things that can go wrong, first of all, the lack of semantic clarity about what is the words that we're using mean. And this is particularly a problem as we move into digitizing paper records. We still have billions of paper records in the industry that need to be digitized. But they also have to make sense, and they have to make sense in a correct fashion. And the terminology that we use is tended to be very lax and very informal. And truly the way each individual discipline, each individual region thinks about what a term means can be really different. What a well is in Canada is not the same thing as what a well is in the United States, which is not the same thing as well as in Brazil, as is defined legally in regulations.

So, these things cause what we call in PPDM dissonance differences in the way people manage their information. And we try to develop some best practices and standards that help us to reconcile the problems that we have with dissonance to help us to find and correct errors in data. So, we can find links that are in Faraday's and change them to feet or whatever is right to get all those dates reconciled into the same format.

Just, you know, dates are really complicated problem. We just published an 18-page white paper explaining some of the main kinds of problems with dates that we find in data management in industry. If you don't think that there's 18 pages worth, everybody might go have a look at our website and have a read of it and send me an email and tell me what you think of it. But dates are really an interesting challenge.

We also see in spreadsheets and paper documents, lots of content or context of data, the attenuation. And that is what you saw today that we're striving to pull not just the written content out of each document, but to get the contextual representation, the metadata around that that gives the data that you're capturing its meaning. And that's really important too. But we have to remember every company does their data, thinks of data in a slightly different way. And try to reconcile those differences between organizations is one of the focal points of the members at PPDM.

So just very briefly, the things we do will work on semantic disambiguation’s, which I talked to you a little bit about. We have a library of standardized reference lists that that you can pick up and using your own implementation. They're well defined, so it helps us to resolve some of those problems with vocabularies. We have a large library of over 6,500 data rules for verifying quality of data and strong semantic languages, data object definitions, et cetera. So, if you're interested in seeing how industry can collaborate to try to make this transition to AI and other technologies easier for everybody, I'd invite you to come visit us at www.ppdm.org. And that is that's it for me. So, Dean, turn it back over to you.

DEAN One of the things I'd like to talk about, I just wanted to bring this up because it's current. But I was at the Global Energy show that's going on right now in Calgary. And my premise for being there was to see what some of the companies are doing around AI. And I thought I'd try to bring that to you just to talk about it.

So, you know, one of the companies I seen, they actually had a little bit of a use case where you could take your website, like our company website, they would pass it through a system, their system, compare it to their website, and they'll come back with all the things that maybe are in parallel between you two so that maybe things that you could partner with as a company. And I thought that was that was kind of innovative. So, that was one of the things that I thought was quite neat over there.

Another one is that I saw was they they're using automation, so robotic automation, and it's for welding. And so, they had a couple of devices set up that does the welding, and they were showing this. And they said one, the larger one could take the place of 22 welders. And I was like, oh, my, that that is impressive. But I can't imagine how much that would cost. I think you could pay those 22 welders for a lot of years, but whatever.

You know, I seen a lot of companies that were focused on field operations. So, you know, again, a big area that we mentioned about E&Y there at the start, you know, some people said, you know, we're just we're just going with the wave. AI is it's still new, it's in its infancy. So, so we're just trying to keep up and use it where we can. So, a lot of a lot of companies were using it on the services side of field operations. And so, I thought that was that was quite interesting.

Yeah, and I mean, there was another company that they they do LiDAR imaging. So, you they have a device that you hook to your iPhone, and they will just walk around and capture equipment. So, you could use it by use case would be capturing equipment, and you could tag that equipment. So almost like an asset management type system. So, you know, very efficient it. Yeah, and those were some of the things that I saw. You know, there was about 380 exhibitors there. So, it's a big show at the BMO Center here in Calgary.

And a smaller subset were where these AI focused companies, but that's not really what the conference is all about the conferences more about operations and drilling completions production engineering, but it was really enlightening to see what some of these companies are doing and really quite amazing. So, I just wanted to share that. I'll maybe turn it back to you Eliza.

ELIZA WITH PANDELL Dean Michelle and Trudy, thank you so much. I thought that was really interesting. And it's exciting to see all the changes that are happening in the world, AI driven changes.

ELIZA WITH PANDELL But this is a question for Canadian oil producing company moving from hard copy files to digitizing with hopes of using AI. Is there a place or organization where we can find best practices, legal requirements, etc. And there seem to be many different options and experiences in terms of how to get this job done. So, any recommendations with regards to that? You can all pipe up if you have an opinion.

DEAN I mean, I'll jump in, and you know, some of my Trudy here. Yeah. A lot of people and a lot of good standards around it. That's for sure. So, I'll let her jump in after me. But I will also add that, you know, you know, I don't typically like throwing out names. But then, you know, one of the companies that that we work with a little bit is to get companies out of that paper format is West Canadian. So, they're based in Calgary here. And they've got a 75-year legacy of, you know, digitizing oil and gas companies and other energy companies too. So, Trudy, maybe I don't know if you want to add more substance to that.

TRUDY Well, I agree that there are more companies out there wanting your business than you can shake a stick at. If I was going to go about doing this, and I had a lot of legacy documents that I wanted to have digitized, I'll tell you what I would do, because I'm a very suspicious person. I would have somebody go and find the 20 or 30, the worst documents you can find, things that are not very nicely written, that have crap in the margins, all that kind of stuff. We'll find a bunch of those and then pick your top five AI people and tell them to go do what they can with those five.

And then, first of all, see how long it takes them too, see how many questions they have to ask you. And three, see what they come up with, and determine which ones are actually doing the best jobs. There is nothing that separates the sales pitch from reality quite as much as doing an actual test. And given the level of competition, I think I would do that.

ELIZA WITH PANDELL That seems like a really great suggestion, Trudy. Yeah, I agree.

MICHELLE WITH PANDELL Well, Trudy, I actually, you know, we have a process for that. There are pilot programs that we run for our land record services department. And so, we say, send over some documents, and we'll show you our process, what it can do. And to go back to what we were talking about right before is that, you know, AI isn't 100% accurate, and humans aren't 100% accurate. But if you put the two together, you're going to have a very accurate. And so that's why we always have our land records experts verifying the data that goes through the AI processes. So, it does get better than over time, right? The models get better and smarter, but you always have to have that human in the loop to double check because you don't want it. If you miss it, right, if you miss a decimal, if you miss something, and you know, you can lose the lease, you could, you know, anyway, go ahead, Trudy.

TRUDY So, the credentials of the people doing the work and the credentials of the system are both important. And they're not very true. They could have really talented people in a horrible system, or a terrific system that's run by people who don't really fully understand what they're talking about. Land documents in Canada are very different than land documents in other parts of the world. So, these credentials, the verification of how that engine's been trained, if you've got an AI engine, and it's been trained on American documents, and then you try to turn it loose in a bunch of Canadian documents, it may not have a very good time. So, I think be really careful, do your homework, and make sure they understand how the AI was trained, understand its credentials, and then give it a hard job to do and see how it does. Don't give it easy stuff, a monkey could do the easy stuff. That's a good point. Yep, good point.

ELIZA WITH PANDELL I love it. I love it. And I love the fact that we've received a bunch more questions too. So, do we want to, do we want to fire through them? Okay, woohoo, go team, this is great. Okay, is there a way to measure the ROI that can make the justification more palatable for companies?

DEANThat is a great question.

TRUDY And that is really difficult. What you need to do, but you need to measure before, and you need to measure afterwards, and you need to figure out what it is you're trying to achieve, and then set your metrics, and start capturing the metrics before you do the study. Because otherwise, you have nothing comparative to work on. Yeah, like they're not any more magic than anything else.

DEANAnd I'll just add, you know, my experience for the last 30 years has been around land systems. And, and honestly, I truly do feel that land is kind of the center of the universe. So, I look at it, and is there really anything more important than your land records, right? So, land's always in charge of all the partners, your working interest partners, your royalty partners, your landowner, stakeholder engagement.

So, ROI, yeah, it's, Trudy's exactly right, very challenging to gauge. It really takes a good business case up front to figure that out. But definitely around land, I think it should help alleviate that business case, because it's such an important area for any energy company, in my opinion.

TRUDY Okay, I can push out a cross reference if that's okay. A lot of land systems have in them, because of the nature of the partnerships that are created, and the way data is shared between partners, the entitlements that are laid out in those land documents can impact things like how you manage your technical world data, your technical seismic data, your production data, and if those documents are hidden inside the land system, and they're not made available to other users, you can make actionable errors with your seismic data or your well data. So, there is a liability component that needs to be taken into consideration as well.

ELIZA WITH PANDELL Yeah, good point. Good point. Okay, thank you, Trudy. We'll hop on to the next question, if you're good with that. What challenges limitations do you foresee for land staff with companies wanting to make their own AI models in their network for security concerns versus existing models outside of their network?

DEANI think I'll jump in if you want. So definitely, AI is so new, right? And I don't believe corporate policies are there yet, obviously, like I would think that most of you online, if you talk to your compliance groups, there's no policies around it yet, right?

So, I think the way the AI is going to go, the way the industry is going to go, is they are definitely creating their own proprietary platforms internally that are not connected out to the worldwide web. And I think, you know, eventually, if they do that, they create those models internally, you keep asking it questions building on the model, then it's going to be your proprietary information.

Those policies will have to catch up to say, okay, you can't, a staff member shouldn't be taking my working interest partners or my royalty interest and uploading those and asking a question around Co-Pilot. I think that's where we got to start really being concerned. But you know that opportunity is there, you can do that, right? But I just wouldn't recommend it. But I think those policies, they're going to be coming out, I'm guessing in the next year or so. I think we're going to see companies releasing those. Trudy, Michelle, what are your thoughts?

TRUDY I'm filled with horror at the very thought.

MICHELLE No, I have nothing to add to that one.

TRUDY It's maybe no worse than our legacy practice of taking all those royalty interests and partner interests, you mentioned them, and taking a copy of them and then replicating them in some other system and then failing to update them when the partnerships change. So, you know, the problem is equally heinous. We just need good data governance and practice to avoid those and stop people from making impulsive decisions that are bad.

ELIZA WITH PANDELL Good advice. Okay, we've got two more questions. This one's directed to Trudy next. From a seasoned data professional perspective, what general advice would you give to energy companies, especially the ITIS and legal functions teams, that are digitizing land records into structured data?

TRUDY That's a great question. Land records, because they're really individualistic, are a challenge to integrate and harmonize. It really is a challenge to share with other organizations, because the language you use and the terminology that you use is grounded in the culture and the education of the people who are creating those documents, and in the legal framework within which they're generated.

And more and more, these documents that have huge amounts of inferred content in them are being shared with people to use in analysis, etc., who don't have the expertise to understand those nuances. And that means that, I mean, it means a lot of things, but one thing it means is that we have got to start teaching people critical thinking again, because taking things at face value, just because it was on Fox News or whatever, doesn't mean it's not true, or that it is true, right?

And we need to provide environments where people can quickly learn what they need to learn or understand the pitfalls that they're going to fall into. I think there are people who think that putting everything into an AI will obviate the need for humans, and maybe someday it will. But that will only happen when we have achieved a level of semantic consistency globally that allows us to do that. Right now, the variations are simply too high, the risk is very high, that if things are going to be misinterpreted. So, I would proceed with an attitude towards knowledge and syntax and semantic retention. It's not just AI, if that makes sense.

ELIZA WITH PANDELL Okay, thank you. All right, we'll hop on to our last question thus far, which is, given the current pace of innovation and adoption, would you consider this to be early days of artificial intelligence development?

MICHELLE I'm gonna go with yes. Trudy, you go first.

TRUDY AI is still drooling in the crib.

[Laughter]

ELIZA WITH PANDELL All right, say no more. All right.

MICHELLE Go ahead. Do you want to say something else?

TRUDY No, that was it.

MICHELLE That was it. That was it.

TRUDY It's a toddler soon. [Laughter] Don't assume too much. I mean, it's super cool. Come on.

MICHELLE Yeah, super cool, but super early. And so, I have a technical team here at Pandell that I work with every morning, and they are the brains behind our AI and land records. And every day, they're coming up with something new or something to try in the AI or something. So even though it's early, we're rapidly learning. And so, who knows where the limit is of AI. So yes, to answer the question, we are new early days. It's really in a crib.

ELIZA WITH PANDELL So yeah, that's a good way to say it.

DEAN I'm stealing your thunder, Michelle. But one of the use cases that Michelle and her team worked on fairly recently is they get a bunch of PDFs in from a project, right? And there was thousands of PDFs. I'm talking a couple tens of thousands of PDFs.

And they noticed right away that some of these documents were missing the first page. Well, the first page is very important because that's your indexing page. But that kind of tells the basics of…

MICHELLE …grantee dates, you know, the legal description. Yep, yep, location.

DEAN And so, it was kind of scary because, you know, the team had based kind of the cost estimate and everything around that first page and what it would take to extraculate everything. And so, we had some people that Michelle was working with here, and they ran it through AI, thousands of documents that came back within a couple of minutes and said these 5,300 and some documents are missing the first page. So, you either have to go back to the source, or you have to plan out a different scope.

And so, you know, to have someone, an individual go through that would have taken forever to figure that out. But those are the, you know, AI is never gonna, it's not going to replace people, in my opinion, it's just going to allow people to be more efficient, and to do things that are more important, like you're not dealing with that, that low level stuff anymore.

MICHELLE So, I take care of the tedious routine tasks, or yeah, the easy possible due diligence projects. Yeah, yeah.

ELIZA WITH PANDELL So, for that reason, we should embrace it, right? We do have one more question, maybe. While going the way of digitized records, especially when there's the possibility of assistance from AI is such a dream, are you aware of any land records that must be retained, must be retained in their physical format? This person has gone through a few companies now where the lawyers have advised that any document granting an interest in land like leases needs to be retained in their physical format, but not all companies have this policy. What do you think? That may be one for Michelle.

MICHELLE Yes, I see that P. Dubb, I see you on here. I see he has an answer, he answered that question for his. Oh, he did too. Yeah, yeah. Try to retain all physical documents as our contingency plan in case of database outage or cyber attacks.

And that's really, it depends on the laws around it. So, I visited a utility company not too long ago, and its law in that state to keep their records forver. So, I think it really depends on whether what the law is, or what the company's policy is in preference.

I mean, I've been scanning my own personal documents for years now and I have none, and I haven't lost any of them yet. I mean, they haven't had a corrupted hard drive. And that's just me, but I know that corporations, they have better penalties in place. So, it's really up to, yeah. Yeah. I think it's really important that I'm sure to, you know. Trudy, do you have anything to add to that? Go for it.

TRUDY The legislation in every country is evolving, and it's really important, and you have to take into account not only whether or not you can replace a paper document with a digital copy of it, and sometimes that depends whether it's got a wet signature on it, or in some cases, if it's handwritten.

So, in Alberta for a long time, you were not allowed to ever destroy the handwritten survey notes for field surveyors, because that was considered to be the definitive records. But digital acts are being passed now that allow those to be superseded under certain circumstances. But if you're going to do that, you also have to take into account any laws that have to do with data sovereignty or data residency.

So if you can't just in every region take your documents that are legally, need to be legally admissible in evidence, you have to be careful where those cloud environments are, and you have to check for data sovereignty and data residency legislation, because if you do it wrong, you can inadvertently render your documents inadmissible in evidence.

So, I think that's the thing we always have to consider is what does the law say, and what is the impact going to be if something bad happens and we get sued, and we've got to put those documents in front of the judge.

ELIZA WITH PANDELL That's a good point, and a good side thought. Yeah, we may not always be considering that side of things, but that's a really valid point for sure.

Okay, well I think that was a really exciting conversation, and we did get through everybody's questions. So, thank you, team, you did great. And thank you to everybody for joining us online today. It was a really great audience, nice big group of people, delighted to have you join us and hope you'll join us with the next one.

We love doing these for you, and we love the community we're building by doing it. And yeah, hope to see you again. Thank you so much Dean, Michelle, and Trudy. It was great to have you online today, great to get your expertise, and I hope everybody has a wonderful afternoon.