WEBVTT 00:00.000 --> 00:08.400 Let's start with the next session, and the last session of the day, which is a lighting 00:08.400 --> 00:15.000 talk by Yordi, you see throughout your vets, a business intelligence architecture for social 00:15.000 --> 00:17.000 and solidarity economy. 00:17.000 --> 00:18.000 So, I would say you. 00:18.000 --> 00:20.000 Thanks very much. 00:20.000 --> 00:29.000 Good evening, first nightly session, the last and maybe the list. 00:29.000 --> 00:32.000 But, go on. 00:32.000 --> 00:37.000 I'm going to talk about business intelligence architecture for social and solidarity economy. 00:37.000 --> 00:39.000 I'm Jordi Seljuvet. 00:39.000 --> 00:41.000 I'm the tensioner. 00:41.000 --> 00:46.000 In my city, I put the tensioner and business intelligence and security issues. 00:46.000 --> 00:47.000 But, it's a lie. 00:47.000 --> 00:49.000 I'm a SQL-Holigan. 00:49.000 --> 00:50.000 I'm working in co-depth. 00:50.000 --> 00:53.000 It's a worker's cooperative. 00:53.000 --> 00:56.000 We call a self-resover for social economy. 00:56.000 --> 01:03.000 We only use free software, we only work for social and solidarity economy entities. 01:03.000 --> 01:06.000 But, what social and solidarity economy is? 01:06.000 --> 01:11.000 Social and solidarity economy refers to a wide range of economic activities that tend to 01:11.000 --> 01:17.000 prioritize social and profitability instead of poorly financial profits. 01:17.000 --> 01:24.000 NGOs, cooperatives, associations, foundations, and all type of entities that 01:24.000 --> 01:29.000 want to transform the world to become a better place. 01:29.000 --> 01:36.000 And, of course, social and solidarity economy companies like other companies need to make decisions based on data. 01:36.000 --> 01:37.000 Sometimes they don't know. 01:37.000 --> 01:39.000 Sometimes they don't want to know. 01:39.000 --> 01:41.000 Sometimes they uncomfortable with that. 01:41.000 --> 01:44.000 But, they need it. 01:44.000 --> 01:49.000 When you want to develop business intelligence project from social and solidarity economy, 01:49.000 --> 01:54.000 first of all, you need to analyze how we social and solidarity economy. 01:54.000 --> 01:55.000 Okay. 01:55.000 --> 02:00.000 You can find car sharing co-operative, telecommunications co-operative. 02:00.000 --> 02:10.000 Lawyers or bulls, architecture bulls, marketplaces, supermarkets and a lot of different kinds of stores. 02:10.000 --> 02:14.000 Schools and a lot and a lot of different businesses. 02:15.000 --> 02:21.000 Sometimes you can find many mature IT departments and databases with thousands of millions of 02:21.000 --> 02:22.000 rows. 02:22.000 --> 02:26.000 Sometimes you can find databases with hundreds of thousands of rows. 02:26.000 --> 02:30.000 And sometimes you find it with rows. 02:30.000 --> 02:35.000 Sometimes you can find entities with only one data source, traditionally an ERP. 02:35.000 --> 02:40.000 Usually you find entities with a lot of different data sources. 02:40.000 --> 02:44.000 Sometimes you find entities with a low budget. 02:44.000 --> 02:47.000 Usually you find entities with a very low budget. 02:47.000 --> 02:48.000 Don't worry. 02:48.000 --> 02:52.000 You never find entities with a high budget. 02:52.000 --> 02:59.000 And of course, you have to develop business intelligence project on any kind of project for social and solidarity economy. 02:59.000 --> 03:01.000 You must use visual. 03:01.000 --> 03:05.000 But not for price because sometimes Google or Microsoft offers free tires for NGOs. 03:06.000 --> 03:15.000 The importance is for building community for sharing knowledge for protecting user privacy in short for freedom. 03:15.000 --> 03:20.000 We have seen that we have a very different kinds of entities in social and solidarity economy. 03:20.000 --> 03:23.000 We have different needs and different architectures. 03:23.000 --> 03:28.000 And also they have a very low budget and they want to interoperate. 03:28.000 --> 03:32.000 Then we need to build different architectures. 03:32.000 --> 03:37.000 And we build different architectures for each type of entity. 03:37.000 --> 03:40.000 Smalls, mediums, large, or asteroids. 03:40.000 --> 03:44.000 Its architecture is based on the previous level. 03:44.000 --> 03:50.000 Then you can start as a small entity and then install the new components. 03:50.000 --> 03:53.000 If you become a medium company. 03:53.000 --> 03:56.000 All architectures are modular. 03:56.000 --> 03:59.000 You can quit from element and install other. 03:59.000 --> 04:03.000 For example, you don't want dbt plus alpha for ETL. 04:03.000 --> 04:06.000 You can install Apache Hop, for example. 04:06.000 --> 04:13.000 And if you, and all entities, all architectures are based on the previous one. 04:13.000 --> 04:17.000 If you become, you start as a small company and then you become a medium. 04:17.000 --> 04:22.000 You only have to install the new components, not the other parts. 04:22.000 --> 04:23.000 Okay. 04:23.000 --> 04:25.000 Hop, hop, hop, hop, we don't it. 04:25.000 --> 04:26.000 We use Ansible. 04:26.000 --> 04:30.000 Because Ansible allows us to choose what to deploy in in this case. 04:30.000 --> 04:32.000 With very eyeballs, with tags. 04:32.000 --> 04:39.000 You can choose if you want to install the progress and the dbt and the superset or whatever you need. 04:39.000 --> 04:46.000 If you are a bigger, you can install a staging area at the lake and so on. 04:46.000 --> 04:49.000 It allows to have to different environments. 04:49.000 --> 04:51.000 The tech changes in software and company relations. 04:51.000 --> 04:57.000 If you define a new variable in your Ansible, then you will install the new components. 04:57.000 --> 05:01.000 If you start a medium company and then you become a large company. 05:01.000 --> 05:08.000 You define the new variables and then Ansible detects these changes and install the new components. 05:08.000 --> 05:09.000 Then you need. 05:09.000 --> 05:13.000 And also, else to encourage variables. 05:13.000 --> 05:16.000 And what we have done from comp depths. 05:16.000 --> 05:19.000 We have an Ansible provisioning of this architecture. 05:19.000 --> 05:21.000 With all the variables, with all the tags. 05:21.000 --> 05:25.000 In order that you can test, you can share in your own communities. 05:25.000 --> 05:30.000 You can do merge requests, you can invest and whatever you want. 05:30.000 --> 05:32.000 With this architecture. 05:32.000 --> 05:36.000 In order to provide all such and such a utility economy. 05:36.000 --> 05:40.000 A architecture for the business intelligence projects. 05:40.000 --> 05:42.000 And also we have Ansible inventories. 05:42.000 --> 05:45.000 With variables defined for our customers. 05:45.000 --> 05:47.000 All are in clear because Ansible and Ansible are in clear variables. 05:47.000 --> 05:53.000 But if you want to test it, you only have to change the capabilities for your own databases. 05:53.000 --> 05:55.000 And deploy on your own servers. 05:55.000 --> 06:00.000 And then you can test the architectures. 06:00.000 --> 06:02.000 Okay. 06:02.000 --> 06:04.000 Let's see some architecture components. 06:04.000 --> 06:07.000 We are going to see the data or housing. 06:07.000 --> 06:10.000 The ETL process and the database relationship. 06:11.000 --> 06:14.000 For data or housing, we use podcasts. 06:14.000 --> 06:18.000 Because it's free because it's robust and has a good community. 06:18.000 --> 06:19.000 And also has extensions. 06:19.000 --> 06:21.000 I like a lot to extensions. 06:21.000 --> 06:23.000 One is time scale. 06:23.000 --> 06:27.000 Not all of us to write very efficient queries from time series. 06:27.000 --> 06:30.000 That it's the most used case in data or housing. 06:30.000 --> 06:33.000 And the other is data for internet or wrapper. 06:33.000 --> 06:37.000 For internet or wrapper, all of us as to access different data sources. 06:37.000 --> 06:40.000 As if they were in the same database. 06:40.000 --> 06:43.000 You can cross different sources. 06:43.000 --> 06:45.000 For example, CSV approaches database. 06:45.000 --> 06:49.000 And my SQL and other projects database in only one query. 06:49.000 --> 06:53.000 And it's very nice for first thing to market. 06:53.000 --> 07:01.000 If you have small companies or million companies with little data. 07:01.000 --> 07:05.000 Because you can cross data very easily from all the sources. 07:05.000 --> 07:10.000 And then put it in a superset and basically. 07:10.000 --> 07:13.000 It allows to simulate the E of an ETL process. 07:13.000 --> 07:16.000 Because it's similar that you have the data on your own server. 07:16.000 --> 07:18.000 But it's not on your own server. 07:18.000 --> 07:23.000 Then if you have a big databases, you will have a big effort with that performance. 07:23.000 --> 07:26.000 How do you solve that with the ETL? 07:26.000 --> 07:32.000 We solve with the ETL because it also does it transform a lot of an ETL system. 07:32.000 --> 07:34.000 And it's SQL in had to be changed. 07:34.000 --> 07:39.000 And for SQL, it's very nice because you can do ETLs only with SQL. 07:39.000 --> 07:43.000 And also, it assumes that an ETL documentation automatically. 07:43.000 --> 07:44.000 I have two things in the world. 07:44.000 --> 07:46.000 Races people are working in documentation. 07:46.000 --> 07:49.000 And the details of me the second one, partially. 07:49.000 --> 07:53.000 Intatso has time control, data quality, and so on. 07:53.000 --> 07:56.000 In green, data sources. 07:56.000 --> 08:01.000 Blue, the medium pass, in orange, the final table. 08:01.000 --> 08:06.000 All tables with the columns, with the time column automatically, 08:06.000 --> 08:08.000 without doing anything. 08:08.000 --> 08:13.000 We can see in the previous talk that you can add more information on your models, 08:13.000 --> 08:16.000 with the description, the description of its case. 08:16.000 --> 08:25.000 But automatically, makes this only with SQL plus little ginger and some more information. 08:25.000 --> 08:29.000 ETLs on the ETL lineage, only with SQL. 08:29.000 --> 08:32.000 Finally, for data visualization, we use superset. 08:32.000 --> 08:35.000 We have seen three cults previous. 08:35.000 --> 08:42.000 It's an iterative session for BWR, with SQL plus ginger, if you want. 08:42.000 --> 08:46.000 And with filters, it's a standard session, user control, 08:46.000 --> 08:47.000 really, really security. 08:47.000 --> 08:49.000 That's very powerful. 08:49.000 --> 08:52.000 This is the powerful feature. 08:52.000 --> 08:55.000 Embring graphics in all websites, set service charts. 08:55.000 --> 08:59.000 Your users can make your charts with track and drop. 08:59.000 --> 09:01.000 That's very nice. 09:01.000 --> 09:05.000 We can see, as well, you can change the colors. 09:05.000 --> 09:07.000 I'm very bad choosing the colors. 09:07.000 --> 09:10.000 But you can do graph, choosing the type, and then, 09:10.000 --> 09:12.000 that can drop for data sets. 09:12.000 --> 09:15.000 That's very nice. 09:15.000 --> 09:19.000 Finally, I want to emphasize the importance of free software 09:19.000 --> 09:21.000 for social and solidarity economy. 09:21.000 --> 09:24.000 Because it allows to have all types of projects, 09:24.000 --> 09:28.000 but specifically, business intelligence projects and systems, 09:28.000 --> 09:32.000 and strengthening the community, protecting the user privacy, 09:32.000 --> 09:34.000 and helping the intercombration. 09:34.000 --> 09:39.000 Thanks very much to everyone to develop, to support, to use free software. 09:39.000 --> 09:42.000 Please don't stop doing that. 09:42.000 --> 09:47.000 It's very important for changing the world and changing the economy. 09:47.000 --> 09:52.000 You want to test our architecture or enhance our architecture. 09:52.000 --> 09:56.000 We are very happy to accept full request, 09:56.000 --> 10:00.000 and share on your own communities, 10:00.000 --> 10:04.000 and things like match, that's all. 10:04.000 --> 10:09.000 Thank you.