Name: Automate Your Way Out of FOI and Public Records Requests Headaches
Uploaded: 2025-05-08T14:16:04.360Z
Duration: 45 min
Description: Automate Your Way Out of FOI and Public Records Requests Headaches

Transcript for "Automate Your Way Out of FOI and Public Records Requests Headaches": Hello, and welcome everybody to this webinar about automating your way out of FOI, Freedom of Information, and Public Records request at Aegis. Today, we will, show you in this webinar how you can efficiently handle your freedom of information requests and your or your public records requests. A quick introduction first on the, right hand side and on your screen, you will see a q and a section. That's where you can enter your questions. And after this webinar at the end of the webinar, we will try to answer all your questions in this session. Quick introduction about us. My name is, Daniel Schroedink. I'm responsible for product management within Reveal for certain of our products. And with me today is Carl Martens. Carl, can you introduce yourself as well? Yes. Thank you, Daniel. I'm Karl Martens, also colleague of Daniel. I'm involved in advising and implementing our services in governmental processes, typically here in The Netherlands where it all started. And so we have central government, state government, local government, and government agencies of all types. And they, many of those use our services for the FOIA requests, part of their inquiries, but more on that later on in the presentation. Yes. Thank you, Carl. Yes. I think we have been working with government agencies now for over eight to ten years in this area, helping them with, handling their FOIA request or FOI request. I think we have, quite a few different names within Europe concerning this. I think in The UK, we call it freedom of information. In, The US, it's called freedom of information act. In The Netherlands, we say the. I think in Germany, it's the information. So quite a few names that are available for, you know, let's say, this aspect. And basically, in every European country, as a citizen, you are allowed to request data from your government. And by law, many countries by the law that many countries have, the government will release the data to you. Now there's a lot of variation within Europe, on how that is done and what the exact requirements are for We see some countries very much a focus also on the redaction of black lining of documents. In other countries, we see more a focus on the workflow and the, time to deliver the data to the citizens. So there are quite a few, different scenarios that we also want to address today. We want to show you today how with the software that we have available, how we can help you make that full process of releasing documents to the public, making that more efficient. A few words about Reveal. Reveal developed software. We developed software for, investigation of information, and we call it the discovery solution. But that solution is also ideal to be used within, FOIA of MOI and public records request, because, basically, the process is kind of the same. You will need to, find the relevant data. You will need to do some filtering and selection of the data that you want to release or have to release to the public. You need to refill your data, redacted data, and in the end, publish your data in a proper format for the public to be able to read that information. Now it's kind of a similar process that we do within eDiscovery, and Review has developed some tools to help you with that. Just to give you an overview of some of the clients that are currently leveraging the solutions that review offers, quite a few. I think we have a few from, listed here from The Netherlands, some provinces, local government, like the city of Amsterdam, but also, more, in the central government, like the Ministry of Justice. Garo, can you tell a little bit more about Ministry of Justice? Because I know that you are having a lot of contact with them. Yeah. Yeah. Yeah. My expertise indeed, very much focused on Dutch government and then central government, state government, for VINCI's, as we call it, in The Netherlands, and then city governments and government agencies. Typically, for The Netherlands, we have a process of actively publishing public, governmental data, but still our what we call the passive, publication requests, they focus on what we call the FOIA. And and that's very much involving email email from previous ministers, state councils, directors of services. So a lot of email is involved, so the volumes are high. And that's happening at the state governments. That's happening on the central governmental level. So, we have a few Dutch events which track attention from journalists and also law firms and and the public in general. If you look at the COVID, for instance, or we have the the liquid gas, discussions here in The Netherlands, which caused earthquakes, they all the the the evidence needed for the public and for the journalists is always via these FOIA requests. So they are large and and and huge. So if I estimate a little bit, we have on daily the daily frequency is about 1,000 active users every day in in average, from governmental side working on the files there, which are uploaded in our cloud, to to do the, the FOIA process. More more on that later via via Daniel. Yeah. Thanks, Carl. Yeah. And one of the other names you see here is Disla. It's, the child and family, agency in Ireland leveraging, the software for, public records requests. Basically, that was a result of, one of the large investigations in Ireland, the mothers and baby commission investigation about the, childcare, from the past, which had a huge impact. So, yeah, quite a few clients that we have in this space, throughout Europe. If we take a look at the the, the products that we feel offers, we've quite a nice lineup of products, which are not all relevant for the Freedom of Information solution that we here offer. But just to go through them, we have tools or products to collect data from various sources. So that means that you can extract data from your own bill systems, from, chat systems like Teams and Slack, and all kinds of other systems that we support over 25 systems to be able to get data easily out of systems so people don't have to export that manually and put it into a system, to process the data, but, the collection of data can be fully automated. I think legal hold is not really relevant in this case. It's more a US oriented product, to, fulfill some requirements in the the law there. Today, we will focus on Logical, which is a very simple tool to help you with your request, and we will also do a demo of that today. Review is a more advanced solution, leveraging a lot of AI in the product and used for, let's say, the very extremely large cases as well. And then we have a tool called trial director, which allows you to present your data in court. But today's focus will be, Logical, and Logical is a web based solution that helps you with your collection of data, the review of your data, redaction of your data, and publishing it into a unified format. A little bit more about Logical. Logical was designed with, yeah, let's say, the user in mind, to make it really simple to use advanced technology for to help you analyze your data and process your data. So it's very simple to get your data into the system. After getting the data into the system, we will normalize your data, making sure that you can search all through all the data, without having to do any complex processing or settings that you need to change. So we really try to we allow you to search and filter for data as much as possible, to identify quickly the information that is requested. And after that, there are all kinds of solutions to help you, for example, to detect automatically personal identifiable information and help you redact that information. Now we'll take a look at the FOI, challenges, yeah, to help you to publish the proper data that is being requested. Yeah, we have seen we know from experience that the data can be anywhere. Although many organizations have, let's say, document management systems where data resides that is sometimes requested. We have seen that most of the requested data that we see is buried somewhere in email boxes or in chat environments because that's where the, let's say, important data is on why decisions, for example, has been made. So, the big challenge one of the big challenges is to get all that data into the system and make that simple. And, also, in several countries and wars that it this process cannot take too long, but this tool will help you to quickly identify and get data into the system. Another challenge is the, the various formats that are out there. We have, already talked about chats, emails, but there are also PDFs, PowerPoints, Excel files, all sorts of formats that needs to be reviewed. You don't wanna use different products for that. We do everything in one single application. We also seen, that there are always a lot of duplicate documents, and I know we have some numbers on that from, some of our clients. And I think, Gara, wasn't it, like, sometimes up to 80% of the documents that are collected and gathered for a request seem to have, seem to be duplicate. Yeah. Typically, for request, of course, there's some kind of topic. And then, the best practice we know is that, mailboxes are backups from people in in public service, people involved in that. And when you have eight to 10 people, which which from which we upload these mailboxes, about 70 to 80% is duplicate, which reduces the data a lot, then reducing not only the data, but reducing the time needed to process all that and also, to do to reduce the risk in disclosing, protected data, for instance. So deduplication is very important in the process. Yeah. And, also, I think, typically, what you mentioned now is that data resides in mailboxes. And the typical workflow we have is that we just collect data from the mailboxes. Also, in those mailboxes, a lot lot of non relevant data is there, like, newsletters, sometimes even personal data, and it's really important to be able to identify that quickly and get it out of the, let's say, the set of documents that you want to publish in the end. Yeah. I think one of the things which, we see a lot is that handling FOI requests, public records request always involves or in many countries, involves also redacting personal identifiable information in order to protect civil servants. We see that this process can be quite cumbersome. You see here a couple of examples in the picture that we have. These are actually released documents. The top one is one that's still black lined with a marker in a very, very bad way. But, apparently, still organizations use this technique, and, yeah, you can yeah. I have the suspicion this is actually based maybe even on purpose that this data is leaked. But, yeah, this is, of course, not the proper way to Hey, Daniel. If I may add here, this is the link to this Ministry of Justice here on The Netherlands, and they were preparing a, FOIA process. And then they were printing out emails to to hand them over to the law firm who would then scan them in again, put them into Adobe to do the, the redactions, And they saw, of course, they saw the technology fit in how we could do it directly, import the mailboxes, select relevant documents, deduplicate what's in there in duplicate as a duplicate, and then start the redactions automatically, which was a very big time win. And, of course, very important for them is the time to deliver the documents to the requester is very important and was optimized very much by using this technology. Yeah. I think we still come across some organizations that still print email and then scan it in again and then use Adobe to, to redact or even use markers in some cases. And then print email, take a black marker, and then, copy it a few times just to make sure that everything is done correctly, which is kind of, let's say, tasks you don't would not like to do. Yeah. And the the example below is also an issue with, which we come across sometimes, that data is leaked because the redaction was the doc prokarylene. This is an example of the AstraZeneca contract that was released by the European Commission, where in the document itself, everything was nicely redacted, but, people forgot to also redact the the bookmarks, which actually contained the text. So redaction is quite a big issue. Of course, there are more things. For example, if you want to review email and check that, it's important to be able to see instantly also all the related emails so that you have a full overview of the, of all the documents in the email thread or in the conversation. Access to the attachments in the emails is really important. So you have a proper overview. Indeed, identifying PII, doing the redactions, yeah, quite a few things which are important, which we will address later in the, in the demonstration. Now in order to help you with this process, we've designed a workflow which we call, yeah, our best practice, and we actually use this in, yeah, that's I think almost 95% of the clients are using Yeah. The workflow, as we suggested. But, basically, the steps are we try to keep it it's a simple process. So first of all, you collect and upload your data that you think is part of the request. After that, what we do is that we filter and filter the the data, for example, for data that is not relevant. For example, data that is, coming from newsletters. Very simple trick is just to search in your data for unsubscribe because then you will find all the email newsletters that are being sent to, people, and you can just take those out. You can assign documents to the people, your reviewers, your team, and after that, you will start with your review. If possible, you instantly take a decision to, publish the documents, and release the document, do a full release or a partial release, you can have a QA over quality control, on the document and then produce the document as we call it, publish the document, including a, in UK, I think they call it a schedule, but, same inventory list of all the documents that you publish with the numbering, the names, and the exemptions that you have in your document list. Of course, you can also take decision that the document is out of scope, so we put it aside. But, also, in some cases, the information you need to review can be quite complex. And then you can also send, this to a case expert or, domain expert to, review the document and take the decision and bring it back into the process to, for the quality control and reduction. Daniel, if I may add to that, so a case expert is one, that's internal. Of course, you can also address external third party, which is happening in here in The Netherlands very much. So third parties also have to give their, review to specific documents when their interest is can be harmed by the request. And it's also good to mention that, we have a legal tech team sitting here in the office to help you online in, for instance, these filtering queries. And so you have a request, which is, of course, natural language, and it has to be transformed into some kind of a, Boolean query, so a search query. We are always happy to help, customers in creating those queries because those queries very often are also shared with the people who have done the request. So they know exactly how we search, what is searched, and where it is searched. Okay. Yeah. Thank you. So what I would like to do now is, give you a demo of the solution that we offer. So what I will do, I will share with you my screen of, our logical application. And what you see here is a web based application that allows you to support the full process that we just showed, uploading data, reviewing the data, and publishing your data. Well, normally, what we do is that we create what we call a new project for every request that is coming in. That can be easily done by, the persons that have the rights to do this. I already created a project for a public disclosure case in The UK. I will open that project. The first page where you go to is that you go to what we call the upload section where you can upload your data. I already have done a few uploads here. You see the various uploads. So there's already some data in this project. But it's basically the first stage is that you're going to collect your data. Now there are a couple of ways how you can collect your data and get your data into the system. One of the things that you can do is do a upload from, for example, a cloud application like Slack or maybe from, Microsoft three six five. Then you can select mailboxes or Teams, to get your data straight into the system. So there's no need for an export. If you have the right to access certain mailboxes, then you can get those that information straight into our system. Another way to do this is do a file upload, and I'll just give you a quick demo of that. I can drag and drop files and folders in here. Also drag and drop directly from Outlook if I want to do that. But in this case, I will just upload some files from my computer. So I'll take, let's say, a few minutes and other documents that I collected. And what will happen now is that the spouse will be, selected, and I will give this a upload name. So this is a demo. One thing that I need to add is a custodian. So where's the data coming from? Now the custodian is, let's say, the person who's legally responsible for the data. So I'm the owner of my mailbox and the owner of my laptop. So I'm will be, in this case, the, custodian. I'll start the upload, and what happens now is that this data is being uploaded to the system. And after the upload, the system will, what you call, will start processing all your data. And what happens then is that we will analyze your data, try to detect PII, but also extract attachments. If you have uploaded the zip file, we'll extract the contents of that. So there's nothing you really need to do yourself. That whole process of, extracting and making your information ready to review is done automatically. So this takes a while, but what we can do in the meantime, we can go to the search section. And this is basically where you will have access to all your documents that you have uploaded. Now you see here, in this case, I only have 125 documents, but I can also have 10,000, hundred thousand of documents in my system depending on how much data I collect. You see the number of pages. And on top of this, there's an option to search for you through your data. Below that, we have a few filters because what we do is that we try to extract a lot of metadata from the documents. So for example, you will see here that we define also the document types. So if I want to, for example, focus on emails or other types of, let's say, images or spreadsheets, let's say in this case, email, I can just select one of these filters that you see here. And now I will only have here in my recent list the emails. So it's a very quick way of, selecting documents. Another option that I have, once I filter for email, for example, that I can filter also for email from or email to. So let's see, all the email that has been sent to me and kinda just quickly tick these boxes, and you see that, again, the result list will change, and the documents will be, selected here. Once I've got the result list and I've got here a few documents, then I can, for example, assign all these documents and, like we mentioned. So I can select all on this page or all in my search, and I can assign it to a specific user, for example, Carl in this case. And I will assign these documents to him so that he can start reviewing these documents. So all these documents have now been assigned to Carl, and he can go to his list and review these. In order to view documents, you see here the result list with some information, and I can just double click here this document. And then you will see that the document is being opened, and I can start reviewing this. On the right hand side, you see that we have some text available, which are part of our workflow. That means that we can start labeling documents. So for example, we want to take a decision on this document if we want to do a full release, partial release, or maybe it's privileged or we want to withhold this document for a specific reason, or we can say, well, this document is out of scope. So, for example, if you say, now this document is out of scope, we just simply select out of scope, and the document will have a label out of scope. And now we can go quickly through the next document, and we can review this and mark this, for example, to perform a full release or a partial release. Now one of the things you see in the document is that there are some lines here, some curly lines, the orange lines, which indicates that there is PII in the documents. So that's all automatically done. So in this case, I already see PII, and I can say, well, this will be then, for example, partial release because we will redact the entity information. I can do the redaction right now if I want to do that, And then you will see here that I've got here some PI detected addresses, email addresses, and there's an option to redact instantly that information. You can add here a redaction label, and these are some labels from The UK, and label this as being personal information. Let's say redacted PII. You will see that all of the information is now instantly redacted. So no need to drag, boxes on the data. System detected the PII. Only one click, you can redact all that. Of course, you can also do some manual redactions. So if you want to redact some information as well, we can do it here and then change, for example, the, subscription or, add your own add your own description here. So you can go really quickly through these documents, save your changes, and go through all of that. So let me go back to the initial page. So you also see now that the label that we added out of scope, partial release are added to these documents. In addition, what you can do is that you can also search for data. So let's search, for example, for a simple search for plan. And if I search here, let me clear first my initial search. So I'm back at the starting point, search for plan, and you will see that I found here some documents. Of course, I can also combine this or change this to plans. So let's see here the plans here. You'll see that you can quickly navigate through the documents. So adding a search, executing the search basically instantly. Of course, once you've done the search, you can also use the filters here. For example, search for the email or search for other things. So let me open here this document plans. And what you see instantly here is that this document, if you take a look at the family of the document, also has attachments. It actually has images. So if I click here on the attachments, you will instantly see here, for example, the attached, plans that were created. So and this is really important that you have, let's say, this this family as well so that you can easily navigate through all the items in an email and also go through, let's say, documents that are similar or documents that are duplicates. Those are all shown here. Now once you have gone through all your documents, that you wanted to review, what you can then do is that you can and you have done the reductions and you have added, the proper labels. You want to basically export all your data to publish your data to the requester. Now for that, we have what we call a download option, and, the downloads is quite simple. You create a new download, and in this case, I already created a template, a simple template for downloading the data. That template defines how I want to publish my data. It defines in which format I publish the data. It defines the metadata I want to publish, like in UK, the schedule, but or the inventory lists, as some other companies call it. And then we'll do that based on the search. So, I can select one of my searches. So, for example, ready for production. Those are two documents in this case. So this is a search that I created before and saved. So these are the documents that are ready. And if I go through this wizard, it will show me, okay, something which we call a load file. And a load file is this the schedule that we create. It's a Excel file. It sees people not with a few properties from the documents, which you can also define yourself. Like, for example, the reduction codes, which are important. So the reduction reasons, the exemption codes, those are part of the template. A couple settings how you want to produce, so we want to get everything out as a PDF in multipage. We will not share documents in their native format in this section, and, also, we want to include the redactions. A couple of things what you can do. Let's say you want to withhold a document and document has a withhold tag, then we will not publish that document, but instead we will create what we call a slip sheet with information that this document is being withheld. And then your schedule or inventory list can contain, the reasons for that. So you also create a name for your download. There's a summary here, but I think more important, you can also share your download easily with the recipient. So there's an option to securely share your this download So you don't have to send things by email or do other things. Just simply click and it won't securely share, enter a email address, and the recipient will receive a download link. And you can also decide how long you want to have that download week the the download link available and when it should expire. So that's a nice option. And after you've done that, and then you scan here, create the download. I already did the download here, so, let me show you the results of this download. Let me open up here the file. So what you're basically getting is that you get a folder with, images as we call them, and those are the documents that we have published. And this is one of the documents that you will see. And this is one of the documents that we, redacted, and it will show it on screen like this. So it will be nicely redacted document. Besides that, what we also do is that in the data section, we will store here, for example, an Excel sheet, and this Excel sheet will contain all the the proper metadata that you can use, in columns. And then the data, can be used to for a company document that you, publish with the actual documents. So that's in short. And, of course, there's a lot more to tell and to show, how you can review quickly your documents and publish them. You can use this for large datasets, smaller datasets, but it is a simple and single tool to support your full workflow. So it's not only a reduction tool, but it is a full featured tool that allows you to collect data from various sources, to process the data, search through your data for the relevant documents that are applicable for the request, and then, publish that information. If it's good to add that we are all happy to provide a one on one demo online for people who would look would seem like to see more details on how specifically the work can be performed. Yeah. Definitely. Because I think it's always good to see it by yourself, and we can even do it with your data. So if you supplies with some of your data, which is public, we can actually show to you how we'll work with with your data. Yeah. So just to summarize what I've just shown, the collection, the filtering and assignment, assigning the documents to Cargill, doing the first pass review, making a decision, selecting, in this case, partial release, performing the redaction, and then do a production so that the data can be published. I think, yeah, the options for case experts, you just label it to be assigned to a case expert, and then that person will need to review and go through the same process. Now all this together, yeah, some of the challenges we went through are listed here again. If you have to do everything manually, it can be slow and labor intensive, especially in the reductions, but also filtering, reading through all the documents. There can be a lot of data. Also think about the deduplication, which can be a huge time saver if you're collecting, you know, let's say, too many documents. Redaction, helping the AI, helping you to identify personal identifiable information. And, of course, you will probably always read the full documents, but it will help you and automate parts of that. And, of course, maybe it's important to understand we have, we ordered all the actions in the system. So there's a full transparency of what happens to the data. I think, yeah, one of the things to take away as well is the, that you can collaborate with the team on this. Everybody can access the same projects, not only one person, but you can work with multiple people together in a project to speed up your process. And, yeah, I think the logical solution is quite easy to learn. You can quickly start with it as with this as well. Basically, you can start tomorrow, collect your data, put it into the system, and maybe by the end of the week have, finished your, your review of your documents and publish that. Yeah. I think it's now time for our q and a section. So just as a reminder, on the right hand side, you can add your questions if you have any questions. And also important, on the top of the screen, you will see a button called Reveal demo. If you press that button, we will reach out to you for a demonstration. So you only have to click that to get that, to be contacted by us. So let's move to the questions. Let me open up this. And I think, yes, I can share the question on screen with you. So, one of the questions is is, if you can change the workflow or for reviewing the documents. Yes. So depending, let's say, on the workflow you want to use and probably also depending on which country you're you are in, you can define, the way you want to label your documents, take your documents. You can also change the the naming of the tags, to match the language that is used within your country for these requests. So you can make it as flexible as you want and suited to your needs. Another question that I see. How many users can access the system at the same time? Yeah. So you can collaborate with, as many people as you like on your, public record request. So whether you have one or two people in your team or five, six to 10, we also have teams with up to 70 lawyers that are working in the same system to, review the documents, redact them, label them. So, yeah, you can ask have as many users as you want. We also have seen in other types of applications even more users with, over 303 viewers or even more. So you can have quite an extensive number of users that will be able to use the system. Yeah. Also a good question. How many documents can be added for a request? Of course, every system has its limitation, but I don't think that, you're running to any limitations of our system. You can add up to hundreds of thousands of documents, or even millions of documents to the system. So and, yeah, I know from experience that most requests and the data that is collected for these requests are, yeah, not, that large, maybe 10,000, maybe up to a hundred thousand documents. But, there are, of course, always exceptional cases. So you might have maybe millions of documents, but that's possible to add it to the system as well. And just as a reference, we have one client, which is, yeah, I think one of the the larger client clients we have in this space that has up to, or even over 15,000,000 documents in one project. So that's quite a lot and should yeah. I don't think many organizations will have more than that. And then I see, I think, one more question. And this is about the detection, I think, of personal identifiable information. So the system allows you to or can detect personal identifiable information. And what we mean by that is that those that's information that is linked to a person. So it's gonna be a person name, it can be an email address, it can be a phone number, it can also be an address, but also Social Security numbers are part of that information. So in most cases, it is all the information that you would like to redact because of all kinds of, privacy regulations. So there's quite an extensive list of, information that is automatically detected and which can be redacted with one click. So I think those were the questions. So, everybody, thank you again for joining our session today. And, yeah, let's see. If you want to have a demo, click on the top right button, review demo, and we will be in touch with you to, provide you with more information. Thank you.