[00:02.080 --> 00:03.320]  My name is Khalil.
[00:03.440 --> 00:07.240]  I'm a software developer at the Organized Crime and Corruption Reporting Project.
[00:08.940 --> 00:13.560]  I develop applications for strategic civil society.
[00:13.680 --> 00:14.600]  That's how I would term it.
[00:14.700 --> 00:18.120]  I try to work with the good guys on investigations and digital security.
[00:19.500 --> 00:24.160]  There is a lot of overlap with the hacker and maker spaces here.
[00:24.400 --> 00:28.100]  And actually walking around, I see that I do have a lot of mutual friends with you folks.
[00:30.620 --> 00:31.280]  Azana.
[00:31.480 --> 00:32.060]  Oh, yeah.
[00:32.620 --> 00:32.900]  Hello.
[00:34.200 --> 00:35.180]  Hi, everyone.
[00:35.400 --> 00:36.660]  My name is Azana Cheman.
[00:36.980 --> 00:42.380]  And I am the product manager for data platforms at the Organized Crime and Corruption Reporting Project.
[00:42.560 --> 00:45.340]  I joined OCCRP in 2021.
[00:45.520 --> 00:49.300]  I'm about to hit my four year anniversary in just a few weeks.
[00:49.440 --> 00:59.340]  And my role primarily focuses on overseeing the development and enhancement of our investigative tooling at OCCRP, such as OCCRP Aleph.
[00:59.840 --> 01:16.200]  In practice, that means working with different teams to ensure that we are building the tools that journalists need, tools that are easy to use, high in performance and functionality, and that are aligned with our mission of exposing broader corruption and crime across the world.
[01:16.200 --> 01:19.340]  So that's our little introduction spiel.
[01:22.840 --> 01:27.980]  So a little bit about the Organized Crime and Corruption Reporting Project, OCCRP for short.
[01:28.140 --> 01:29.980]  You can try saying that five times quickly.
[01:30.640 --> 01:35.540]  We are one of the largest investigative journalism organizations in the world.
[01:35.540 --> 01:38.480]  We are headquartered in Amsterdam in the Netherlands.
[01:38.520 --> 01:42.660]  We previously used to be headquartered in Sarajevo, Bosnia and Herzegovina.
[01:43.100 --> 01:46.820]  And we have about 125 members on staff.
[01:47.140 --> 02:02.680]  That number is significantly lower this year, as we had to unfortunately lay off a quarter of our workforce earlier this year, as the current administration found it that international development funding should not be given out to organizations such as ours.
[02:03.820 --> 02:07.720]  Other than that, our focus, our staff is primarily editorial.
[02:08.100 --> 02:13.040]  However, we have different teams, such as the product team, the product department, which I'm a part of.
[02:13.220 --> 02:19.540]  We have the tech department, our platforms team, our research and data team, which also encompasses our data desk.
[02:20.580 --> 02:23.520]  Overall, we are a mission-driven organization.
[02:23.540 --> 02:27.500]  We publish stories that expose crime and corruption around the world.
[02:27.680 --> 02:31.880]  And we provide... we are a media development organization at the same time.
[02:31.880 --> 02:37.760]  We provide support to over 75 local media outlets in different countries all around the world.
[02:37.980 --> 02:39.420]  We call them member centers.
[02:39.840 --> 02:45.280]  And essentially, we give them everything they need to collaborate on these cross-border investigations.
[02:45.480 --> 02:53.560]  From the tech, to the infrastructure, to the legal support, physical and digital security, to research and data resources, and more.
[02:54.200 --> 02:57.780]  If you'd like to learn more, you can just visit our website, OCCRP.org.
[02:58.220 --> 03:02.800]  There's also a handy-dandy QR code here, which you probably can't see.
[03:03.960 --> 03:08.840]  But this is just a screenshot of the homepage of our website, OCCRP.org.
[03:08.840 --> 03:14.440]  If you're interested in seeing the investigations that we do and the impact that they're having, please give us a visit.
[03:15.320 --> 03:19.000]  Essentially, on our website, this is where the output of our journalism lives.
[03:19.220 --> 03:20.520]  So, all of our projects are here.
[03:20.660 --> 03:21.740]  You can learn more about us.
[03:21.900 --> 03:24.120]  And you can access our key resources this way.
[03:24.760 --> 03:29.820]  You'll also be able to find a list of these local media outlets that I mentioned and our partners accordingly.
[03:31.920 --> 03:32.520]  Right.
[03:32.920 --> 03:38.840]  So, besides being a newsroom, we are a network, like Azana alluded to, and a resource hub.
[03:39.000 --> 03:47.260]  So, we share resources among all of our member centers, especially tech, which would be hard for any one member center to build and administer themselves.
[03:49.520 --> 03:56.580]  Our flagship application is called Aleph, and this is an investigative data platform for journalists.
[03:58.660 --> 04:06.940]  Unlike a lot of other tools, this is specifically oriented towards investigative journalists, so not law enforcement, not necessarily hackers.
[04:07.340 --> 04:10.980]  It's meant to answer editorial use cases.
[04:11.760 --> 04:16.720]  And it's also meant to be a sort of catch-all platform.
[04:17.840 --> 04:31.900]  I can say, when I was a developer, before I was working in media, just working with investigative nonprofits, I found that journalists were really awful clients because they wanted everything like right now.
[04:32.360 --> 04:36.140]  And it was impossible to get through a full development cycle with them.
[04:38.000 --> 04:41.860]  And they also didn't necessarily know what they wanted when they started a story.
[04:42.640 --> 04:50.960]  And I think when you see journalists get attached to developers to tackle a particular story, they don't usually get very far.
[04:51.800 --> 04:54.840]  The solution to a situation like that is a platform.
[04:54.840 --> 05:00.520]  So, a general solution to anything that could feasibly come up in the course of an investigation.
[05:00.600 --> 05:07.840]  So, you want the data to already be there, ingested and organized in a way that you would expect to find it when you query.
[05:10.940 --> 05:22.080]  So, specifically, a user would be a journalist who logs in and is able to search across all of our data in all of our jurisdictions of public records and including their own documents as well.
[05:24.460 --> 05:25.980]  So, just a few of the features here.
[05:26.100 --> 05:27.460]  We're going to do a deep dive on these later.
[05:27.620 --> 05:32.100]  But you can upload your own material and extract information from it.
[05:32.200 --> 05:39.160]  You can create network diagrams and also cross-reference data across different domains.
[05:42.140 --> 05:44.140]  So, why was Aleph created?
[05:44.180 --> 05:46.680]  To make our lives easier or harder, right, Clel?
[05:47.240 --> 05:47.920]  Just kidding.
[05:48.440 --> 05:54.180]  But the real reason that we created Aleph was because journalists were getting all of this data.
[05:54.460 --> 05:57.400]  Leaks, registries, court filings, etc.
[05:57.760 --> 06:00.300]  And half of the time, they didn't know what to do with it.
[06:00.540 --> 06:02.300]  And they didn't even know where to put it.
[06:02.440 --> 06:04.620]  So, we had to find a solution to this.
[06:04.740 --> 06:12.360]  It's almost like solving a giant puzzle where all of the pieces are scattered all over the world in different countries, languages, file formats and more.
[06:12.360 --> 06:15.440]  So, we tried to make a one-stop shop of sorts.
[06:16.000 --> 06:20.660]  So, Aleph is named after a short story by Jorge Luis Borges.
[06:20.840 --> 06:27.600]  And it was developed as part of OCCRP's broader mission to support transparency and accountability in reporting.
[06:27.600 --> 06:34.820]  It centralizes all of that data from those corporate registries, leaks, legal filings, financial records, and more.
[06:35.060 --> 06:42.740]  So that journalists can actually search, cross-reference, and make sense of it instead of drowning in this data chaos, as we like to call it.
[06:43.520 --> 06:48.940]  When it comes to Aleph, we're not in the business of just hosting or housing data, rather.
[06:49.340 --> 06:52.100]  We are looking for a place to make it usable.
[06:52.580 --> 07:05.200]  So, with Aleph, as Clil mentioned, reporters can upload their own documents, you can connect them to other data sets in the database, and you can securely share your exclusive material with other reporters during your active investigations.
[07:06.040 --> 07:13.880]  Everything that we've listed, all of these functionalities, all are key in exposing cross-border financial crimes, organized crime, and high-level corruption.
[07:17.410 --> 07:21.590]  So, I wanted to give a few examples of the investigations powered by Aleph.
[07:21.590 --> 07:30.470]  It is fair to say that any project or any investigation that you'll see on our website was powered or has had some sort of touch with Aleph along the way.
[07:31.230 --> 07:34.650]  So, the first one I'd like to highlight is just our Swiss Secrets project.
[07:35.050 --> 07:39.670]  This project exposed hidden Swiss bank accounts used by criminals and corrupt officials.
[07:40.250 --> 07:43.890]  Reporters on our team, and this was a cross-partner investigation.
[07:43.890 --> 07:46.510]  There were hundreds of people working on this investigation.
[07:46.970 --> 07:57.350]  Reporters used Aleph to analyze this leaked banking data and cross-reference it with public records that we made accessible, revealing just how much that money was sheltered behind Swiss secrecy.
[07:58.990 --> 08:00.970]  Next, we have Dubai Unlocked.
[08:01.210 --> 08:09.670]  So, this investigation revealed how Dubai's booming real estate market has become a safe haven for corrupt politicians and criminals to stash their wealth.
[08:10.410 --> 08:16.910]  Aleph's cross-referencing tools allowed reporters to connect those property records with individuals across different data sets.
[08:18.550 --> 08:20.730]  Next up, we have Scam Empire.
[08:21.250 --> 08:25.310]  Here, there were multiple related leak batches that were ingested in Aleph at speed.
[08:25.870 --> 08:34.750]  Reporters also used Aleph's API to enrich that data and even search through millions of audio files and hundreds of thousands of video files by keyword.
[08:36.210 --> 08:44.270]  In our Dear Compatriots investigation, journalists leveraged Aleph's API to translate a massive Russian email leak.
[08:44.430 --> 08:53.210]  This enabled multilingual collaboration across newsrooms worldwide, making it possible for reporters in those different countries to connect the dots together.
[08:55.350 --> 09:00.050]  With our Bankera investigation, we saw Aleph's flexibility in action here.
[09:00.270 --> 09:06.810]  So, we were able to structure and connect all of these highly complex data sets in Aleph to go deeper into the data.
[09:07.090 --> 09:09.670]  And finally, we have the steward files.
[09:09.910 --> 09:20.770]  Another example of how Aleph was central in helping reporters sift through and organize large volumes of leaked material, revealing financial mismanagement and abuse of public trust.
[09:20.770 --> 09:38.810]  So, whether it's banking leads or leaks rather, banking leaks, real estate records, email troves, or anything that comes our way, Aleph keeps proving itself time and time again as a core tool in the investigative journalist space that helps our colleagues turn all of this data into impactful
[09:38.810 --> 09:39.470]  investigations.
[09:42.630 --> 09:43.190]  All right.
[09:43.390 --> 09:44.390]  So, those are the results.
[09:44.630 --> 09:46.290]  Those are the stories that come out of Aleph.
[09:46.570 --> 09:57.430]  And we want to, in this talk, because a lot of you are technical people, to talk about some of the behind the scenes and pop the hood and to mix metaphors.
[09:57.630 --> 09:59.570]  This is what we call the data iceberg.
[09:59.750 --> 10:02.230]  This is taken from my colleague Eric's previous talk.
[10:03.670 --> 10:09.670]  The idea behind the data iceberg is that there are certain types of journalism technology that you'll see as a consumer of news.
[10:09.870 --> 10:11.890]  And there's a lot that goes on behind the scenes.
[10:12.290 --> 10:15.370]  So, if you're opening a story, you might see the data visualization.
[10:15.430 --> 10:19.130]  You might even get exposed to some of the analysis that went into the data.
[10:21.030 --> 10:29.170]  But what forms the basis of the story are issues like, how do you even know that something is happening and worth writing about?
[10:29.270 --> 10:32.910]  How did you know that this company was even registered to this person?
[10:33.270 --> 10:37.690]  How did you know that they had an asset declaration that left out that company?
[10:39.490 --> 10:41.430]  How did you even acquire the data?
[10:42.370 --> 10:43.850]  How did you get it ahead of time?
[10:43.970 --> 10:48.330]  And how did you have it in a way that was searchable to you exactly when you needed it?
[10:49.390 --> 10:54.630]  So, these are the kinds of questions that our engineers are making applications to support.
[10:54.970 --> 10:59.870]  This is what most of the work of Aleph is engaged with.
[11:02.550 --> 11:07.650]  When it comes to what kind of data do journalists need, the short answer is all of it.
[11:07.990 --> 11:09.470]  Whatever they can get, we'll get.
[11:09.610 --> 11:11.270]  And we're more than happy to take it.
[11:11.410 --> 11:12.110]  So, please share.
[11:12.110 --> 11:20.070]  But overall, Aleph brings together both structured data, so databases and spreadsheets, anything tabular really.
[11:20.410 --> 11:25.930]  And unstructured data, such as scanned documents and PDFs, which we get a whole bunch a lot of.
[11:26.550 --> 11:31.650]  At this point, Aleph contains over 4.4 billion records or entities.
[11:31.650 --> 11:38.550]  And we run around... we run about 150 scrapers on a regular basis to keep our public data fresh.
[11:39.430 --> 11:46.250]  In addition, Aleph houses some of the world's biggest leaks, biggest troves from banking data to email leaks and so on.
[11:46.650 --> 11:57.610]  And our journalists, we give this... we make this accessible for our journalists so that they can move forward and analyze it and specifically cross-reference it across other data, which we'll show in a demo very shortly.
[11:58.350 --> 12:07.830]  On top of that, just as we've mentioned before, reporters can upload their own documents to their own secure workspaces, share them with their friends who are on Aleph, their colleagues.
[12:08.730 --> 12:18.550]  And it's the combination of having all of these records, all of these entities, all in one place with the exact functionality that we provide that makes this tool so powerful.
[12:21.850 --> 12:24.590]  Aleph Ali Cannotti, Alright, let's get a little bit deeper.
[12:26.410 --> 12:26.930]  Exactly.
[12:28.430 --> 12:35.030]  Just to frame the technical challenge that we are trying to solve, as Ana mentioned, that broad category of data.
[12:35.210 --> 12:39.910]  So we need to, first of all, collect all of that in a sustainable, easy to maintain way.
[12:40.750 --> 12:43.370]  And that could mean crawling websites.
[12:43.370 --> 12:55.370]  It could mean maintaining scrapers that can work through web browsers or maybe getting around proxies, things like that.
[12:56.010 --> 12:59.250]  We need to standardize all of that data across those sources.
[13:00.170 --> 13:06.230]  And what that means is, we'll talk more about this, mapping them onto a common semantic model.
[13:07.810 --> 13:18.850]  We want to do kind of the same process on unstructured data, so we want to also get semantic data from text, from images, from PDFs.
[13:20.550 --> 13:32.190]  And we want to put all that together in a really, really, really big search index that spans all of these sources and languages and also has some sort of basis for cross-referencing them.
[13:33.830 --> 13:41.650]  The data base logic needs to be structured in a way that allows for, I guess, robust update logic.
[13:42.690 --> 13:48.630]  Journalists are always going to come with questions about where data came from, how long we've had it for.
[13:49.130 --> 13:50.730]  They're very interested in the lineage.
[13:52.230 --> 13:58.510]  And we, of course, want to wrap this into a web application and serve it to our users according to their permissions.
[13:59.590 --> 14:04.930]  I'll add as another requirement, this needs to be very scalable and robust and secure.
[14:06.010 --> 14:12.330]  We don't know if tomorrow there's going to be a 10 terabyte leak from some militia that we need to process.
[14:12.330 --> 14:14.310]  We need to just be ready to accept it.
[14:14.710 --> 14:20.150]  And, of course, our users are also the people that are uploading and ingesting their data.
[14:22.750 --> 14:28.330]  So, all that together, I'll start talking about data modeling.
[14:28.610 --> 14:34.890]  So, if I'm going to a company registry and getting data from there that is going to be used in future investigations.
[14:35.270 --> 14:41.730]  And at the same time, I'm going to court records of mentions of companies and people.
[14:43.890 --> 14:47.170]  Looking at this data, you can kind of reason that there's overlap.
[14:47.410 --> 14:48.030]  You know, there's people.
[14:48.090 --> 14:49.490]  People tend to have names.
[14:50.290 --> 14:53.590]  They don't necessarily need to be named in a particular context.
[14:55.690 --> 14:59.330]  But you could make that statement about any person.
[14:59.550 --> 15:03.190]  So, we take that and we map them onto a model that we call Follow the Money.
[15:03.190 --> 15:05.250]  So, it's an open-source ontology.
[15:05.250 --> 15:08.310]  If you're familiar with that term, it's a taxonomy of concepts.
[15:11.750 --> 15:20.310]  And we want to be able to make structured and semi-structured records outlining how they relate to one another.
[15:21.730 --> 15:26.570]  I'll show an example of Follow the Money, I think, as of a previous release.
[15:27.150 --> 15:29.250]  This is our taxonomy.
[15:29.510 --> 15:33.870]  So, you have, you know, persons that are a type of legal entity, which is a type of thing.
[15:34.110 --> 15:42.310]  Of course, they inherit properties in the same way that you would expect if you're familiar with some semantics or some programming languages.
[15:45.030 --> 15:47.450]  I'll talk briefly about properties and statements.
[15:49.930 --> 15:53.510]  So, we mentioned, you know, a person would have a name and a registration number.
[15:53.510 --> 15:56.670]  A company could also have a name and a registration number.
[15:58.830 --> 16:04.770]  These values, together with an entity, would form a statement.
[16:04.970 --> 16:12.230]  So, if you're familiar with RDF or triple stores, this would be basically semantic data.
[16:16.070 --> 16:19.070]  So, we collect that and we put it into...
[16:19.910 --> 16:27.510]  These are just examples of how we might extract data from some text.
[16:27.950 --> 16:35.850]  So, you have an entity, a referent, a relationship, and then another entity that it relates to.
[16:46.990 --> 16:55.090]  So, once you have this data, you're able to put it into a search index.
[16:55.610 --> 17:07.610]  Actually, the more that I work on investigative applications, the more I just think that it's the work of creating really, really weird and out there search indexes to answer different kinds of queries that a journalist might have.
[17:10.070 --> 17:16.650]  So, definitionally, an index would be a mapping of a term or a keyword to a record.
[17:17.650 --> 17:19.610]  You might hear it called an inverted index.
[17:19.950 --> 17:22.370]  It's really referencing the same thing.
[17:24.510 --> 17:27.450]  This is really important in creating a user experience.
[17:27.450 --> 17:31.530]  It also has some editorial considerations as well.
[17:31.690 --> 17:40.270]  So, if you've ever Googled yourself and maybe found something you didn't like or maybe something that was in the first search result instead of the tenth.
[17:40.470 --> 17:46.670]  So, maybe your fifth grade talent show is the first result and your PhD dissertation is the tenth.
[17:47.430 --> 17:50.550]  That would be an example of maybe editorial values breaking down.
[17:51.070 --> 18:02.430]  So, we want... even though this is a decision that's made in engineering context, we try to curate our search index to be fair across all contexts in different languages.
[18:03.510 --> 18:05.430]  So, I gave two examples here.
[18:05.850 --> 18:10.950]  Three, where two are a person named John that's spelled differently.
[18:11.190 --> 18:13.370]  But, of course, the name John would map to both of those.
[18:13.370 --> 18:20.070]  So, this is just an example of small analysis that we do on the way when we index data just to make it work for journalists.
[18:23.170 --> 18:27.230]  The data model makes it easy to update the index when data has changed.
[18:27.510 --> 18:29.830]  So, I alluded to this before when I talked about properties.
[18:30.190 --> 18:37.410]  But, basically, to identify any particular entity, we need a little bit of flexibility in the journalistic context.
[18:37.970 --> 18:42.830]  In fact, even governments sometimes have a hard time conclusively identifying someone.
[18:43.670 --> 18:50.650]  So, you might say that your ID number and your date of birth are identifying features and maybe your name is not.
[18:50.790 --> 18:51.510]  You can change your name.
[18:52.290 --> 18:57.590]  So, our data model allows us to basically nominate certain properties as being identifying.
[18:57.870 --> 19:06.830]  So, if I check tomorrow whose company is registered in a particular state, I don't end up creating a duplicate.
[19:07.090 --> 19:11.030]  I will actually update the previous record based on those identifying properties.
[19:11.930 --> 19:19.950]  This is done with a lot of flexibility to actually allow data journalists to make the right kinds of editorial statements.
[19:22.530 --> 19:26.230]  So, onto the fun part, how to get started with Aleph.
[19:26.490 --> 19:35.790]  So, for anyone in our space, if you're interested, if you are a journalist, an activist, a researcher, an academic, we have a list of categories that are available on our website.
[19:35.790 --> 19:41.510]  You just need to navigate to our service portal at requests.occrp.org.
[19:41.550 --> 19:46.550]  There is a form there that says, Aleph, create a new account or request a new account.
[19:48.030 --> 19:48.670]  And...
[19:48.670 --> 19:49.250]  Question?
[19:49.250 --> 19:50.550]  I just have an easy question.
[19:50.730 --> 19:50.890]  Yeah.
[19:51.110 --> 19:51.930]  Do you work with any...
[19:55.020 --> 19:55.740]  No, actually.
[19:55.860 --> 19:56.140]  We don't.
[19:56.400 --> 19:56.940]  Not yet.
[19:58.440 --> 19:59.120]  Hopefully soon.
[19:59.860 --> 20:00.220]  Hopefully.
[20:00.460 --> 20:01.240]  That's a good word.
[20:01.480 --> 20:02.320]  That's a good question.
[20:02.600 --> 20:09.440]  I asked if they work with any Joe magazine associated with this conference.
[20:10.220 --> 20:10.580]  Yeah.
[20:10.720 --> 20:11.260]  Please sign up.
[20:12.140 --> 20:12.540]  All right.
[20:12.740 --> 20:15.650]  Thank you.
[20:15.770 --> 20:16.330]  Hey, friends.
[20:16.410 --> 20:16.710]  I'm sorry.
[20:16.830 --> 20:19.990]  Can we hold questions for the end to make sure that they have enough time to finish their presentation?
[20:21.270 --> 20:21.770]  It's okay.
[20:21.930 --> 20:22.510]  Thank you so much.
[20:24.070 --> 20:25.770]  So, you can go to our service portal.
[20:25.950 --> 20:26.770]  You navigate to it.
[20:26.910 --> 20:27.770]  You fill out the form.
[20:27.930 --> 20:32.270]  That form comes to me and my team and we review them on a rolling regular basis.
[20:32.290 --> 20:34.230]  We will get back to you with next steps.
[20:34.510 --> 20:36.970]  And that is how you get your Aleph account set up at the moment.
[20:38.050 --> 20:40.510]  Aleph in its current state is open-source software.
[20:40.770 --> 20:44.290]  So, anybody can go ahead and use the technical documentation.
[20:44.350 --> 20:49.030]  If you want to spin up your own instance, customize it as you like, you are free to do that.
[20:49.030 --> 20:58.910]  Our Aleph public instance, which we will be demoing in two minutes, basically, is our public instance where all of our data for our journalistic investigations live.
[20:59.870 --> 21:04.670]  And overall, just the goal is that we made this tool accessible for the public good.
[21:04.670 --> 21:06.470]  We have our intentions with it.
[21:06.530 --> 21:07.970]  It is for investigative work.
[21:08.030 --> 21:10.410]  And we've seen the beauty of it.
[21:11.970 --> 21:12.650]  All right.
[21:12.810 --> 21:15.190]  So, sometimes we'll do this demo.
[21:15.190 --> 21:23.670]  We'll show it to journalists and they'll say, you know, that's great that you have this incredible, you know, investigative application that I can search anyone anywhere.
[21:24.070 --> 21:31.590]  But what I have to do right now is, you know, I'm writing a feature about the mayor's dog's bar mitzvah.
[21:31.870 --> 21:34.290]  And do you have anything on the mayor's dog?
[21:34.790 --> 21:38.230]  And we have to say, no, that's not really a priority for us.
[21:38.230 --> 21:39.630]  We can add it to the list.
[21:39.870 --> 21:41.330]  And they say, ah, okay, fine.
[21:43.610 --> 21:48.590]  So, and then maybe they, you know, they kind of put it in the back of their minds.
[21:50.110 --> 21:52.410]  The things that you can do with Aleph right now.
[21:52.510 --> 21:58.230]  Number one thing I think you should do when you make an account is to log in and set an alert for whatever...
[21:59.110 --> 22:05.330]  I think most journalists have people that just kind of interest them, that they try to follow, like, that are just kind of in their universe.
[22:06.090 --> 22:09.190]  Get notified when that person appears in new data.
[22:11.510 --> 22:16.370]  So, this uses... if you're technical, this uses Lucene as the search language.
[22:16.890 --> 22:22.250]  So, you can just search for raw text, or you can search on a particular field for a particular type of entity.
[22:22.950 --> 22:24.190]  These are just some examples.
[22:24.890 --> 22:26.490]  So, any search can become an alert.
[22:27.990 --> 22:35.690]  I recommend searching... start broad, because there's a lot of data, and then use the filtering and the facets to narrow that down.
[22:35.830 --> 22:37.490]  Azana will demo that in a moment.
[22:38.750 --> 22:40.950]  Cross-reference across data sets.
[22:41.350 --> 22:52.790]  Sometimes journalists will have some documents or a leak or a data set of their own, and all they want to know is, is anyone from my local city council, let's say, in this leak?
[22:53.970 --> 22:56.350]  And so, that's a good use case for cross-reference.
[22:56.350 --> 23:07.990]  So, you don't need to understand absolutely everything about 10 terabytes of emails leaked from a Guatemalan militia, but you do want to know if someone from a company that you're interested in is mentioned in there.
[23:09.510 --> 23:14.870]  So, you could use the search, or you could use the cross-reference if you want to know the relationship between two data sets.
[23:16.090 --> 23:17.690]  Uploading your own files, of course.
[23:19.110 --> 23:30.250]  If you have a lot of files, or if you have a leak of your own, you should do that through our whistleblowing channels, which we'll outline in a moment.
[23:30.730 --> 23:31.810]  And of course, collaboration.
[23:31.810 --> 23:36.690]  If you are working in a newsroom context, absolutely get your colleagues invited as well.
[23:37.610 --> 23:42.610]  And you can set up a space for yourself to organize your files in your work.
[23:45.490 --> 23:49.330]  So, I'm going to invite Clil to have a seat, because we have a bit of a demo to go through.
[23:50.230 --> 24:01.430]  So, for the purposes of this demo, we wanted to take a real-world example of something that happened two years ago, where we received a breaking news alert on our phones, and we decided to look into it.
[24:02.670 --> 24:04.530]  So, this is what we're looking into.
[24:04.750 --> 24:18.090]  We are investigating an explosion that happened at a cafe in St. Petersburg on April 2nd, 2023, which killed Wladen Tatarsky, real name Maxim Fomin, who was a war correspondent.
[24:18.450 --> 24:23.250]  For the purposes of this demo, I am going to... I crossed it out.
[24:23.370 --> 24:27.850]  I want us to take that second part out, because at that time, we did not know what happened.
[24:28.070 --> 24:33.210]  All we knew was there was a bombing in St. Petersburg, and we were able to find the address where that happened.
[24:35.130 --> 24:36.930]  So, here, what are we working with?
[24:37.070 --> 24:37.830]  We have the address.
[24:38.130 --> 24:39.330]  It is in Cyrillic.
[24:39.410 --> 24:40.730]  I will not attempt to read that.
[24:40.730 --> 24:44.450]  But, we know that the name of the place is street food bar number one.
[24:46.470 --> 24:48.490]  Next, it took us about 30 seconds.
[24:48.490 --> 24:57.770]  We went on a domain lookup website, and we typed in that address, and we were able to find the company owner, and we were able to find their tax ID.
[24:57.770 --> 25:01.150]  So, in Russia, the tax ID is basically the company registration number.
[25:01.430 --> 25:04.710]  And from there, we decided, hey, let's go into Aleph.
[25:05.170 --> 25:10.870]  So, if you just bear with me for a moment, so I get my screen back.
[25:12.810 --> 25:14.390]  We do not see that.
[25:20.420 --> 25:21.400]  Just one second.
[25:27.490 --> 25:29.530]  Alrighty, welcome to Aleph.
[25:29.830 --> 25:32.530]  So, this is Aleph when you log in.
[25:32.690 --> 25:37.730]  It is going to look a little different for everyone once you log in, as we do this based off of security groups.
[25:37.890 --> 25:41.570]  Security groups are dependent on what access to data you have.
[25:41.570 --> 25:43.650]  So, I am using a demo account.
[25:43.810 --> 25:45.830]  It looks a little empty, but the data is there.
[25:46.730 --> 25:50.370]  And going back to our demo, let me just find the slide.
[25:51.250 --> 25:54.390]  Clil, actually, can you read me the tax ID number so I can just type that in?
[25:58.110 --> 25:58.710]  Next slide.
[25:58.990 --> 25:59.410]  There we go.
[26:05.740 --> 26:06.140]  Perfect.
[26:06.440 --> 26:21.480]  So, what we did was, within 30 seconds, our team, this is a real-world example, they went on the domain lookup website, we got the taxpayer ID, came into Aleph, and all I did was navigate to the search bar on the top and type in that taxpayer ID number.
[26:22.020 --> 26:25.120]  I am going to click enter, and Aleph is going to do its magic.
[26:26.660 --> 26:28.660]  And we are getting some search results here.
[26:29.260 --> 26:30.260]  Let me take this out.
[26:30.840 --> 26:32.380]  And we are getting some search results here.
[26:32.700 --> 26:38.080]  So, on the center of our screen, this is our little search results pane.
[26:38.080 --> 26:42.700]  And we know that the company name is Concord, so we are basically on the right track.
[26:43.520 --> 26:51.300]  And to filter this down further, I already have a data set in mind that I know that I have access to, so I want to filter it by that data set.
[26:51.800 --> 26:55.820]  So, to filter, we have these handy-dandy filters on the left-side menu.
[26:56.160 --> 27:05.960]  I am going to click on data set, and I am looking for, let me see, the Russian Unified State Register of Legal Entities.
[27:05.960 --> 27:09.320]  This is your Russian corporate registry, basically, your company registry.
[27:10.300 --> 27:12.480]  And I see I have one search result.
[27:12.820 --> 27:15.420]  When I hover over it, Aleph does this beautiful thing.
[27:15.560 --> 27:19.480]  It gives me the Latin transliteration for anything that is not in Latin English.
[27:19.560 --> 27:23.220]  So, I am able to see, hey, this is the same company, Concord.
[27:23.820 --> 27:25.020]  I am going to click on it.
[27:25.400 --> 27:29.880]  And whenever you click on a result in Aleph, it just expands in that same window.
[27:30.500 --> 27:33.900]  I will expand it further so that we have a full-screen view.
[27:34.420 --> 27:37.520]  And what I can see from here is a few different things.
[27:37.680 --> 27:45.740]  So, to the left, we have some information about this company, information what data set it's found in, who is managing that data, which is our team.
[27:46.600 --> 27:51.240]  At the top, I can see assets and shares, owners, directors, and similar.
[27:51.500 --> 27:52.900]  Similar we'll get into in a bit.
[27:52.900 --> 27:59.060]  But what I'm trying to find out is who owns that cafe where that bombing happened in St. Petersburg.
[27:59.620 --> 28:01.960]  So, I go ahead and I click on owners.
[28:02.120 --> 28:03.920]  And we can see there's two owners.
[28:03.920 --> 28:05.040]  A company entity.
[28:05.040 --> 28:06.760]  So, you can see the little icons.
[28:07.160 --> 28:09.820]  Each of our entity types has a relatable icon to it.
[28:10.400 --> 28:13.180]  I can see that 10% is owned by this company.
[28:13.180 --> 28:16.100]  And then 90% is owned by this person.
[28:16.680 --> 28:17.740]  That looks interesting.
[28:17.940 --> 28:19.560]  So, let me click on that person.
[28:24.400 --> 28:26.720]  And lo and behold, who do we have?
[28:26.880 --> 28:29.360]  We have Evgeny Viktorovich Prigozhin.
[28:29.580 --> 28:43.460]  So, if anyone is unfamiliar with who that person is, that is a former Russian mercenary leader, leader of the Wagner Group, an oligarch, who died in a plane crash bombing, allegedly, situation.
[28:43.840 --> 28:46.380]  And you deal with that information as you wish.
[28:46.380 --> 28:50.460]  But what we just did was, we got a breaking news alert on our phones.
[28:50.640 --> 28:52.100]  There was a bombing in St. Petersburg.
[28:52.580 --> 28:54.820]  Within 30 seconds, we got the taxpayer ID.
[28:55.220 --> 28:59.320]  Within 30 more seconds, we found out, hey, this is the owner of that cafe.
[28:59.700 --> 29:02.740]  So, this is just one real-world example of how Aleph is used.
[29:03.600 --> 29:08.820]  And I'm just going to take you through a few more features of the UI, so that you can see the investigation workspaces and the like.
[29:08.960 --> 29:09.860]  I wish we had more time.
[29:09.980 --> 29:12.120]  This demo usually takes me like an hour and a half.
[29:12.120 --> 29:14.260]  But I will just speed through it.
[29:15.060 --> 29:18.040]  When you log into Aleph, this is your homepage.
[29:18.540 --> 29:22.500]  And depending, like I said, on your security group, you'll see different amounts of data.
[29:23.020 --> 29:29.320]  So, in this security group that I currently have access to, I have access to 502 different datasets.
[29:30.140 --> 29:34.160]  So, if you are a journalist, activist, researcher, academic, please apply.
[29:34.420 --> 29:38.480]  The next slides that we'll go through will show you how to get extended data access.
[29:38.480 --> 29:42.560]  But here I have 502 datasets that I can scroll through.
[29:43.120 --> 29:45.180]  I will filter it by leaks.
[29:45.500 --> 29:47.300]  There is one dataset I'd like to show.
[29:47.660 --> 29:53.320]  So, these are our leaks that are accessible to our security group that we're demoing.
[29:53.880 --> 29:56.520]  And I will filter it further by country.
[29:56.780 --> 29:58.400]  So, I'm going to look into Russia.
[29:59.520 --> 30:01.800]  And from here, let's see.
[30:01.940 --> 30:06.240]  The Russian State TV and Radio Broadcasting Company emails from 2022.
[30:06.780 --> 30:09.800]  This is your dataset landing page, your homepage.
[30:10.320 --> 30:11.920]  And here you have different sorts of widgets.
[30:12.440 --> 30:14.560]  We can see all of the different entity types.
[30:14.740 --> 30:18.940]  There are over 1.7 million emails in this dataset.
[30:19.300 --> 30:22.680]  There's over 350,000 tables, 300,000 images.
[30:23.500 --> 30:27.000]  Whenever we upload or ingest data into Aleph, it is extracting.
[30:27.280 --> 30:30.640]  As you can see, all of these, there's over 2 million email addresses here.
[30:30.640 --> 30:32.820]  We have over 4,000 IBands.
[30:33.440 --> 30:36.340]  Each dataset has a little info box on the right side.
[30:36.460 --> 30:37.260]  So, please read it.
[30:37.820 --> 30:40.780]  We do... it is to your discretion how you use this.
[30:40.900 --> 30:44.280]  So, please use your editorial judgment on how you use leak data, for example.
[30:45.060 --> 30:46.880]  But you'll see some more information here.
[30:47.060 --> 30:56.840]  We actually got this dataset from our colleagues at DDoS, Distributed Denial of Secrets, and our team manages it.
[30:56.980 --> 30:59.660]  You'll also be able to see other information about our datasets.
[31:00.080 --> 31:03.340]  So, if they are regularly updated, you will see that.
[31:03.380 --> 31:06.340]  It'll say monthly, weekly, bi-monthly, and the like.
[31:07.860 --> 31:09.140]  I have to move things forward.
[31:09.540 --> 31:13.060]  So, something I just want to show are investigation workspaces.
[31:13.060 --> 31:23.600]  These are your secure collaborative workspaces, where for each investigation that we have at OCCRP, and our partners and our member centers, each investigation has a workspace.
[31:24.460 --> 31:28.600]  So, for the purposes of this demo, I'm just going to open one that we already have populated.
[31:28.940 --> 31:32.180]  I can go through the entire process after, if anyone is interested.
[31:32.980 --> 31:36.360]  But let me just find one that says, HOPE...
[31:38.320 --> 31:39.860]  It is somewhere in here.
[31:40.200 --> 31:41.380]  You can ignore all these names.
[31:41.380 --> 31:42.720]  These are all of the testings.
[31:43.900 --> 31:44.640]  There we go.
[31:44.780 --> 31:45.800]  HOPE 16, back up.
[31:46.060 --> 31:50.580]  So, we created an investigative workspace that has dummy data in it, pre-populated.
[31:50.780 --> 31:53.600]  The entities are already generated, but I'll still walk you through it.
[31:54.200 --> 31:57.200]  This is your investigative workspace landing page.
[31:57.900 --> 31:59.880]  I will navigate to the documents.
[32:00.420 --> 32:03.680]  So, in this investigative workspace, we have a zip file.
[32:04.040 --> 32:06.260]  Most file formats are accepted into Aleph.
[32:06.380 --> 32:09.600]  It's in our user guide in our documentation, what can and cannot be ingested.
[32:09.600 --> 32:12.000]  But for the most part, a lot can come in here.
[32:12.440 --> 32:14.180]  We have three CSV files.
[32:14.540 --> 32:21.040]  And when I open up this company's CSV file, for example, it'll show you the table that comes with it.
[32:21.680 --> 32:24.760]  From here, you can go ahead and generate your entities.
[32:26.020 --> 32:40.200]  This is already pre-done, but essentially, you're just selecting your objects, and you're selecting their relationships, and then you're selecting your unique keys for all of them, mapping them out across the columns that are in your CSV file, and then you're essentially clicking generate entities.
[32:41.160 --> 32:45.960]  So, these entities were already created a long time ago, and we can see them over here.
[32:46.200 --> 32:51.420]  We have about 60 ownerships, 25 companies, bank accounts, and people.
[32:52.140 --> 32:54.100]  I'll just come into the people, for example.
[32:54.420 --> 32:56.920]  I already added myself here, but Clil, I'm going to add you.
[32:57.920 --> 32:59.640]  Just for the purposes of this demo.
[33:00.660 --> 33:01.020]  Yeah, okay.
[33:02.520 --> 33:03.460]  Clil, you're here.
[33:03.460 --> 33:08.240]  And I just added and generated an entity strictly from this UI.
[33:08.760 --> 33:14.300]  Something that I'm going to show quickly is that within these investigative workspaces, you can create network diagrams.
[33:14.820 --> 33:16.440]  We are very big on visuals.
[33:16.440 --> 33:22.700]  So, this is a very nice network diagram with all of our entities and relationships mapped out.
[33:22.920 --> 33:26.860]  It's very easy to create a new entity on the network diagram.
[33:26.860 --> 33:28.520]  So, I'm just going to find Clil.
[33:28.880 --> 33:31.220]  You should be popping up somewhere if I missed you.
[33:31.220 --> 33:33.920]  Yeah, you popped up somewhere, but I can't see from here.
[33:34.480 --> 33:36.720]  But essentially, this is your network diagram.
[33:36.940 --> 33:41.420]  There's other cool features that we offer, such as timelines and lists to help with the visualizations.
[33:41.840 --> 33:44.200]  And then we have our cross-referencing feature.
[33:44.440 --> 33:53.960]  So, I'm not going to compute the cross-referencing right now, because it is going to take all of my entities and cross-reference the 502 datasets that I have access to with my Aleph account.
[33:54.460 --> 33:57.460]  But essentially, this is what most people use Aleph for.
[33:58.000 --> 34:02.520]  I can see my entities, what I inputted into the system or ingested.
[34:02.780 --> 34:04.920]  I can see a possible match.
[34:05.080 --> 34:06.620]  I can see a confidence score.
[34:06.780 --> 34:08.640]  And this is how we start connecting the dots.
[34:09.120 --> 34:11.240]  So, this was Aleph in a nutshell.
[34:11.540 --> 34:12.400]  Very short demo.
[34:12.620 --> 34:15.620]  I wish I could show you more, but we are close on time.
[34:15.760 --> 34:17.320]  We have a few more slides we want to get through.
[34:17.980 --> 34:19.240]  Please request an account.
[34:19.480 --> 34:19.960]  Join us.
[34:19.960 --> 34:22.300]  I would love to talk more about it after the session.
[34:31.690 --> 34:32.170]  Yes.
[34:32.170 --> 34:35.050]  This is how you can request an account.
[34:35.210 --> 34:38.130]  Just navigate to requests.occrp.org.
[34:41.190 --> 34:44.510]  We have, as Izana pointed out, a lot of data.
[34:44.710 --> 34:46.690]  We have the right to have a lot of data as journalists.
[34:46.690 --> 34:50.710]  We also take the responsibility that comes with that very seriously.
[34:50.710 --> 34:54.410]  So, we make sure it's used for the right journalistic use case.
[34:56.750 --> 35:08.090]  Outside of our member centers, I'll say people that have access or newsrooms that have extended access are usually ones that we have, I would say, close relationships with, which we like to develop.
[35:10.210 --> 35:14.810]  We also, for Aleph in particular, we have a discourse group.
[35:14.810 --> 35:17.670]  So, this course is like an open forum platform.
[35:19.170 --> 35:20.790]  We love hearing from users.
[35:21.510 --> 35:26.490]  Of course, we don't like go into your investigations and your spaces and see what you're doing.
[35:26.510 --> 35:30.450]  So, the feedback that we get is going to be from our users here.
[35:31.650 --> 35:33.370]  Please do ask questions.
[35:33.470 --> 35:37.030]  And of course, we have our contact info as well.
[35:37.130 --> 35:40.530]  Or do you want to say a word first about the future of development?
[35:40.770 --> 35:42.090]  I'll just mention something briefly.
[35:43.270 --> 35:52.530]  So, when it comes to our roadmap and what we're planning on doing with Aleph, we are currently working on building Aleph Pro, which is the next generation implementation of Aleph.
[35:53.090 --> 36:00.950]  The time finally came to take this extremely complex piece of software that was built about a decade ago and completely redo it.
[36:01.110 --> 36:04.410]  So, we rewrote the whole thing from the ground up, basically.
[36:04.810 --> 36:13.790]  And the reason that we're doing this is that we can't just think about setting ourselves up for success for today, but we have to think about where we're going to be in 10 years' time in the future.
[36:14.070 --> 36:24.930]  So, what Aleph Pro will bring is a complete rewrite, as I mentioned, modern modular architecture, faster performance, easier scaling, foundation that's ready for extensions and integrations.
[36:24.930 --> 36:29.130]  If you've used Aleph previously, you know it's a little slow, so we're working on it.
[36:29.450 --> 36:42.730]  We are enhancing our ingestion pipeline, and we're supporting more formats, working on our data models, enabling knowledge graph capabilities, working on changing the entire UI coming soon next year, hopefully.
[36:42.730 --> 36:46.370]  So, it's going to be faster when it comes to searching, smoother usability.
[36:46.730 --> 36:49.250]  We're going to have flexible deployment options with it.
[36:49.470 --> 36:54.630]  Whether it's a SaaS offering, private cloud, on-premise, we're going to help with the migrations.
[36:55.170 --> 36:57.770]  And it's going to have a tiered access model.
[36:57.970 --> 37:01.650]  So, Aleph will always forever and ever be free for nonprofit journalism.
[37:01.650 --> 37:05.770]  It is what we do for our space, for our field, for our colleagues.
[37:05.770 --> 37:13.250]  And we're going to offer additional paid tiers for organizations so that we can sustain ourselves when it comes to everything that we're doing here.
[37:14.170 --> 37:21.670]  And if you are interested in early access or beta participation, same website, same service portal, request.occrp.org.
[37:21.810 --> 37:23.950]  Please fill out the Aleph Pro interest form.
[37:24.190 --> 37:26.890]  I, as a product manager, I need to hear from you.
[37:27.090 --> 37:29.250]  So, please, we would love for you to test it out.
[37:33.080 --> 37:33.780]  What's going on?
[37:35.280 --> 37:48.260]  So, if you have data that you'd like to share, or if you're a whistleblower, we do maintain a SecureDrop instance, SecureDrop Freedom of the Press Foundation's application for secure whistleblowing.
[37:49.400 --> 37:51.880]  You can also reach out via email.
[37:52.180 --> 37:57.340]  All of our journalists have encryption keys available on our website.
[37:58.300 --> 38:02.160]  And if you need guidance on how to do that, you can talk to us afterwards.
[38:02.520 --> 38:05.040]  I will say, don't leak to us.
[38:05.180 --> 38:06.860]  I don't want that.
[38:06.860 --> 38:08.020]  I'm not a journalist.
[38:08.620 --> 38:12.460]  But we can guide you in how to navigate that.
[38:16.240 --> 38:16.980]  Shameless plug.
[38:17.260 --> 38:17.880]  We are hiring.
[38:18.200 --> 38:25.300]  So, in case anyone is interested, we're looking for a head of data, investigative data director, a back-end developer, and front-end developer.
[38:25.500 --> 38:27.880]  Clil, do you want to talk about your back-end developer for a bit?
[38:29.120 --> 38:29.560]  Yeah.
[38:29.780 --> 38:35.760]  Or maybe if you're interested and you have some back-end development experience, feel free to chat with me.
[38:35.840 --> 38:36.680]  I'm hiring for my team.
[38:37.460 --> 38:37.900]  Yes.
[38:38.060 --> 38:40.620]  And I'm looking for a front-end developer for Aleph Pro.
[38:40.880 --> 38:43.640]  So, if you want to come work with us, please talk to us after.
[38:45.520 --> 38:48.480]  And, yep, that is our contact info on the screen.
[38:48.760 --> 38:50.180]  Our signal usernames are there.
[38:50.320 --> 38:51.140]  Feel free to reach out.
[38:51.240 --> 38:52.060]  Our emails are there.
[38:52.340 --> 38:55.280]  Our encryption keys are available on our website as well.
[38:55.280 --> 38:58.000]  So, that's how you...
[38:58.000 --> 38:58.920]  That's Aleph.
[38:59.840 --> 39:00.480]  Cool.
[39:18.270 --> 39:18.950]  Great talk.
[39:19.190 --> 39:19.650]  Great work.
[39:20.170 --> 39:20.930]  Super pivotal.
[39:21.410 --> 39:42.990]  I was wondering if, you know, the organization has considered, you know, taking the financial expertise that the group has and applying it to, you know, some of the people who are suffering under these financial controls.
[39:44.830 --> 40:00.750]  The Paris Protocol is one of the facets in how the State of Israel is, you know, helping genocide the Palestinian people and their inability to, you know, control their own finances.
[40:00.750 --> 40:04.150]  And just with the...
[40:05.750 --> 40:13.030]  It seems like the aim of the group to say, hey, you know, how do powerful people use finance for ill?
[40:14.070 --> 40:18.430]  I'm wondering, you know, just what the organization, you know, would just...
[40:18.430 --> 40:22.590]  Even by just bringing it to light about, you know, this Paris Protocol.
[40:23.490 --> 40:36.790]  Some people have, you know, highlighted this of how Palestinians have no access to their own financial services, hence why they can't build economy and, you know, resist this action.
[40:37.270 --> 40:41.750]  I'm just interested in, yeah, what the organization might be able to do to end the suffering.
[40:43.310 --> 40:44.410]  Thank you for the question.
[40:44.410 --> 40:54.430]  So, when it comes to our editorial work, there's not much that we as representatives from, let's say, the product department or the research and data team can influence along those lines.
[40:54.590 --> 40:59.930]  But that being said, we are 100% going to give this feedback to our editorial department.
[41:00.150 --> 41:05.610]  There is, at any given time, I'm going to say 50 investigations that are in the works.
[41:05.610 --> 41:08.250]  A lot of our work is grant-based, I have to say.
[41:08.250 --> 41:12.350]  So, where we get the funds to operate in, we do.
[41:12.570 --> 41:14.650]  And that's just how our NGO lifestyle is.
[41:14.730 --> 41:17.010]  And I hate to say it, that's just the reality of it.
[41:17.290 --> 41:18.890]  So, I wish that we had people on the ground.
[41:18.990 --> 41:21.630]  I wish that we could cover every single region of the world.
[41:21.670 --> 41:23.190]  And I know that's not the answer you want to hear.
[41:23.350 --> 41:25.090]  And it's not the correct answer either.
[41:25.170 --> 41:30.250]  But I guarantee that I will make sure to bring this feedback to our editorial, our editor-in-chief specifically.
[41:30.910 --> 41:32.430]  Just thinking about like, okay,
[41:38.740 --> 41:39.300]  thanks for the answer.
[41:39.480 --> 41:39.800]  Of course.
[41:40.000 --> 41:42.860]  I can say, you can investigate our coverage.
[41:42.860 --> 41:46.440]  We do have member centers in those regions that are doing great work.
[41:46.620 --> 41:48.540]  A lot of investigations have come out of them.
[41:49.640 --> 41:53.660]  I think like Azana said, I don't try to influence editorial as a software developer.
[41:54.120 --> 41:59.420]  I think our mode of change or what we believe in is just investigative journalism.
[41:59.600 --> 42:01.340]  And that's how we drive things forward.
[42:02.260 --> 42:09.080]  There are, I think, plenty of financial investigations in the Middle East and Palestine and Israel.
[42:10.380 --> 42:11.360]  Yeah, check them out.
[42:11.520 --> 42:12.380]  I think they do really good work.
[42:15.020 --> 42:15.540]  Hi.
[42:15.880 --> 42:17.220]  Thank you for the presentation.
[42:17.480 --> 42:20.380]  It's really a cool product and very important work.
[42:20.820 --> 42:23.800]  My question might be a little bit outside of the scope of Alice.
[42:24.060 --> 42:26.760]  But it's kind of like a natural extension for the next step, right?
[42:26.840 --> 42:36.700]  Because I think Alice has the potential of enabling like individuals or smaller teams to conduct investigation journalism at a professional level or a more solid one.
[42:36.700 --> 42:49.940]  So, but in many of those cases that those smaller teams or individuals, when they try to develop a report, they have to face some kind of pushback from individuals or institutes that empower, right?
[42:49.940 --> 42:51.400]  Most of the important reports do.
[42:52.120 --> 42:57.040]  So, where do they, like, where do those reports go, right?
[42:57.140 --> 43:02.360]  For example, an individual or a small team can have their own website, but that have a very little audience.
[43:02.560 --> 43:07.340]  Or they can go through a larger or more popular channel, but then there's more gatekeepers, right?
[43:07.340 --> 43:16.400]  So, can you talk more about that landscape or what do people usually do in that case or, like, for the potential Alice users?
[43:16.560 --> 43:19.780]  If they're small individuals, what are their options there?
[43:20.060 --> 43:20.400]  Yeah.
[43:20.540 --> 43:20.780]  Thank you.
[43:21.440 --> 43:21.700]  Of course.
[43:21.820 --> 43:24.480]  So, I'll just comment that we are a community.
[43:24.480 --> 43:29.040]  So, OCCRP and our network, it takes a network to fight a network, as we say.
[43:29.040 --> 43:37.220]  If we have, let's say, freelance journalists who find something interesting and they need support, they are more than willing to, more than happy to reach out to us.
[43:37.300 --> 43:38.240]  We are happy to help.
[43:38.600 --> 43:46.300]  So, when it comes to working with smaller newsrooms or just individuals on a case-by-case basis, anyone can reach out to our editorial team at any time.
[43:46.520 --> 43:48.580]  I don't think we've ever said no, particularly.
[43:49.000 --> 43:52.940]  We are also able to provide research services as well.
[43:53.040 --> 43:56.240]  We have an amazing research desk as part of our research and data team.
[43:56.240 --> 44:09.140]  We have this service called ID, Investigative Dashboard, which is a research ticketing system, where our colleagues and our network and even individuals who don't directly work with us but need support can use that service.
[44:09.540 --> 44:14.980]  So, I'm just going to say that, overall, we are happy to help out whenever we can.
[44:15.200 --> 44:16.960]  We love that Aleph is being used.
[44:17.080 --> 44:22.200]  We hope that people, more people use Aleph, and it is a free service that anybody can spin up, really.
[44:22.940 --> 44:23.620]  You had something?
[44:24.300 --> 44:24.820]  Yeah.
[44:24.980 --> 44:33.620]  I think, like Azana said, the answer, how if you are a freelancer or an independent newsroom, how can you get into larger channels?
[44:34.420 --> 44:36.100]  So, Aleph isn't a publishing platform.
[44:36.260 --> 44:40.960]  That is an interesting feature set that we've contemplated.
[44:41.300 --> 44:43.120]  Right now, it's just an investigative tool.
[44:43.120 --> 44:54.420]  But if you're trying to get published or you're trying to get access to more serious tools beyond Aleph or maybe domain expertise that our teams have, absolutely just collaborate with us.
[44:54.580 --> 44:55.360]  So, we're a network.
[44:55.520 --> 45:01.500]  We're used to working with people that are on the ground or people that have a particular interest in a story and kind of augmenting their work.
[45:01.600 --> 45:04.900]  So, that's 100% the format that we like to work in.
[45:05.640 --> 45:07.400]  Let's do a question from this side.
[45:07.400 --> 45:08.280]  I think we...
[45:08.280 --> 45:09.320]  We have two questions from Matrix.
[45:10.080 --> 45:11.060]  Are you guys also waiting?
[45:12.180 --> 45:12.460]  Yeah.
[45:13.140 --> 45:13.380]  All right.
[45:13.440 --> 45:14.280]  Should we do one and one?
[45:14.740 --> 45:15.460]  Is it...
[45:15.460 --> 45:15.660]  Yeah.
[45:15.780 --> 45:17.320]  One from Matrix and one from the crowd?
[45:17.940 --> 45:18.220]  Okay.
[45:18.480 --> 45:18.660]  Sure.
[45:18.660 --> 45:19.380]  All right.
[45:19.520 --> 45:28.380]  So, one question is, for smaller news organizations, investigative organizations, what advice do you have when running your own instance of Aleph?
[45:29.040 --> 45:31.900]  Getting public data sets in FTM format?
[45:32.120 --> 45:35.280]  And also, any advice on any more guided training online?
[45:35.480 --> 45:36.020]  Things like that.
[45:36.900 --> 45:37.160]  Sure.
[45:37.400 --> 45:41.600]  So, I will just instruct our friend there who left this comment.
[45:41.880 --> 45:43.500]  Please use our discourse community.
[45:43.820 --> 45:47.880]  So, that is the way to get in touch directly with the Aleph team, myself and the developers.
[45:48.700 --> 45:50.780]  Once a week, we have Aleph Community Days.
[45:50.780 --> 45:56.220]  These are Thursdays, one-hour blocks that you can book with us to come to us with your issues if you're having...
[45:56.220 --> 45:57.480]  If you can't...
[45:57.480 --> 45:59.740]  If you're having technical difficulties or if you have any questions.
[46:00.160 --> 46:05.660]  So, you can just contact us directly through our discourse community and we'll be happy to help.
[46:08.060 --> 46:08.540]  Hey.
[46:08.880 --> 46:12.580]  So, I have nothing but praise for OCCRP's journalism.
[46:12.580 --> 46:17.100]  I think that you guys do some of the best that I've ever read.
[46:18.660 --> 46:19.540]  I'm not...
[46:19.540 --> 46:21.040]  I don't read every piece of journalism.
[46:21.040 --> 46:24.680]  I'm sure there's something better, but it's really, really good.
[46:26.980 --> 46:28.800]  I have a question about access.
[46:29.400 --> 46:32.900]  So, like smaller journalists, all of that supporting that.
[46:32.900 --> 46:50.640]  But how about organizations that call themselves newsrooms, that have journalists on staff, but that publish stuff that is really, really just meant for tabloids and doesn't really have much of an implication for public well-being or corruption.
[46:50.640 --> 46:52.860]  How do you respond to those kind of requests?
[46:53.160 --> 46:55.360]  And do those people get access?
[46:55.560 --> 46:56.780]  Do they get the same amount of access?
[46:57.140 --> 47:02.220]  How do you protect the innocent people that are in the massive data sets you've collected?
[47:02.640 --> 47:03.000]  Yes.
[47:03.280 --> 47:04.040]  Thank you for the compliment.
[47:04.300 --> 47:06.340]  I will be sure to let everyone know that at work.
[47:07.080 --> 47:11.800]  But when it comes to our extended data access, we have a very rigorous application and vetting process.
[47:12.080 --> 47:13.720]  And this is all run through me and my team.
[47:14.500 --> 47:21.080]  Every single person who requests that an ALIF account be created with extended data access has to fill out a set of questions.
[47:21.240 --> 47:23.280]  And they have to provide links to their stories.
[47:23.500 --> 47:28.140]  They have to provide, let's say, somebody that they know at OCCRP, if possible.
[47:28.340 --> 47:29.800]  So, for everyone in this room, you know me.
[47:29.920 --> 47:30.540]  My name is Azana.
[47:30.960 --> 47:32.320]  So, feel free to put me down.
[47:32.900 --> 47:34.880]  But we take it very seriously.
[47:35.180 --> 47:38.960]  So, we do regular audits where we clean up who has extended data access.
[47:39.280 --> 47:43.120]  We make sure that people need to reapply and renew their extended data access.
[47:43.120 --> 47:45.300]  So, we are on top of it, for sure.
[47:47.100 --> 47:53.260]  Yes, the other question was, will ALIF Pro still remain open-source for open ad hoc groups to self-host?
[47:57.030 --> 48:00.270]  So, the answer with ALIF Pro is we're moving to closed source.
[48:00.530 --> 48:02.090]  And I know how that sounds.
[48:02.290 --> 48:08.490]  But given everything that's happened this year, earlier this year, like I said, we had to say goodbye to 25% of our workforce.
[48:08.490 --> 48:14.730]  And we can say goodbye to any United States funding that we previously had for international development.
[48:15.410 --> 48:21.050]  So, we had to find sustainable revenue streams to keep our team employed and to keep our work alive.
[48:21.290 --> 48:23.070]  So, we want to make ALIF what it can be.
[48:23.250 --> 48:27.270]  Like I said, it will be free forever for non-profit journalism organizations.
[48:28.230 --> 48:32.470]  For organizations that maybe cannot pay the price, you can just talk to us.
[48:32.470 --> 48:33.730]  We are a very friendly bunch.
[48:34.090 --> 48:36.210]  We're also just... this is a work in progress.
[48:36.770 --> 48:38.790]  So, we don't know how this is going to turn out.
[48:38.890 --> 48:39.970]  We're very optimistic, though.
[48:40.110 --> 48:42.710]  So, please just talk to us if finances are an issue.
[48:43.610 --> 48:45.750]  But yeah, that is the story of ALIF Pro.
[48:45.970 --> 48:48.710]  And I'm excited that in a few months, we'll be able to demo it.
[48:49.610 --> 48:56.150]  I'll also add that I think it's worthwhile to differentiate between the data and the platform that the data goes into.
[48:57.130 --> 49:02.510]  The platform specifically, which is what handles the indexing and the cross-reference and the serving of the content.
[49:03.530 --> 49:08.070]  That is something we're having to move into a different revenue model for.
[49:09.230 --> 49:15.230]  The data itself, in the past, we've been very open-handed with data sharing what is permissible to share.
[49:15.610 --> 49:18.350]  And that's something that is going to continue.
[49:20.950 --> 49:22.630]  Hi, thank you for the lecture.
[49:23.870 --> 49:33.730]  I was just listening to the news today and got really concerned about the revisionist take for a lot of our Smithsonian museums.
[49:33.730 --> 49:40.650]  And just the re-revision of history here in America.
[49:40.650 --> 49:48.670]  And I'm very worried as, you know, people are continuously importing that into our K-12 schools.
[49:49.610 --> 49:55.630]  What is... Are you interested in any preservation of what was originally mentioned?
[49:55.810 --> 50:07.330]  Particularly in phrasing and tonal changes that we noticed to reframe the effects of very dark moments in our history, like slavery and such.
[50:07.590 --> 50:14.490]  I know governments might use that tactic in a fascist lens to control populations.
[50:16.970 --> 50:21.830]  So, I want to say, when it comes to the concern, we all share the same concern.
[50:21.990 --> 50:23.910]  Believe me, I am right there with you on that.
[50:24.710 --> 50:28.170]  When it comes to what OCCRP can do in this sense, I...
[50:28.170 --> 50:30.350]  Once again, I wrote this down, or I will write it down, rather.
[50:30.550 --> 50:32.130]  And I will bring it to our editorial team.
[50:32.130 --> 50:34.870]  If there's something that we can do here in the United States, we will.
[50:35.370 --> 50:37.690]  We're a little restricted, as I mentioned, about...
[50:37.690 --> 50:42.270]  There are other organizations that do this type of reporting here, and we have very strict focuses.
[50:42.570 --> 50:46.970]  But I 1000% will be bringing this to our editor-in-chief and to our editorial team.
[50:48.610 --> 50:55.230]  Yeah, I'll also add that just the way we structure our data collection process, we are very good about identifying changes on pages.
[50:55.450 --> 50:58.470]  If it's a page that our editorial is interested in.
[50:59.610 --> 51:04.810]  So we would be able to see when content changes, and only update under those cases.
[51:06.690 --> 51:11.250]  And it's...that problem, I think, more generally, and even in other countries, is something we pay attention to a lot.
[51:12.530 --> 51:12.930]  Okay.
[51:14.110 --> 51:14.450]  All right.
[51:14.630 --> 51:15.190]  Thank you, everyone.
[51:15.350 --> 51:16.130]  I know we went over time.
[51:16.350 --> 51:19.150]  But if you have any questions, Chloe and I will be in the back.
[51:19.410 --> 51:20.170]  Thank you.