 All right, so with that public service announcement out of the way we got several people that always joined us to from From the Militano team. So we got a couple of community members online. I mean melt on a melt on L For those of you that are familiar. It's a it's a like a startup within get lab So I'll let the experts describe what what the talent is. It's it's I got to see a demo from Yanis last month in London. I was like pretty excited So yeah, this I'll let me turn things over to you. I'll let you and the team members introduce yourselves and we'll go from there So go for it. Yeah Yeah, thank you, Ray. So hi. I'm Yanis from the Maldana team with me are also Derek and our from our team and I assume that a lot of people watching this stream Let me serve my screen first of all Okay, so and start The presentation here. I hope that you can check it out Yeah, we can see it. Okay, cool So, yeah, I'm I assume that a lot of people watching this dream have not had the time that sounds to check my dead melt I know in detail. So I'm going to start by introducing you to what we are building describe the problem That we are we are trying to show and go through a quick end-to-end demo of melt on so that you can get a feeling of What we are trying to achieve in the end. I'm going to also discuss how someone can help us out and contribute to melt on so I'm going to Start by presenting you how to use melt ono to extract data from your github groups and projects Load the extracted data to a database Transform the data and analyze the final results Yeah So Let's start with what is melt ono melt ono is github's open source product for the whole data life side All the way from extracting data from API's to run that hook analysis Generating reports and creating dashboards We run melt ono as a startup inside the startup with its own group and projects. It is a separate standalone product Not directly connected to the other github offerings We released last month our version one But we are still at a very early stage and we are iterating really fast through features So what's the status quo? The process I just described is called ELT Extracting loading and transforming data and preparing data to run for data analysis That process is challenging and demands a solid infrastructure People who don't really know much about the data management or data engineering are being forced to Learn how to do complicated things. I bet that anyone who has implemented or is responsible for running such an infrastructure Will agree with me that you have to be really involved throughout the whole process you have to find or implement and then maintain extractors for All the API's that you want to support For example, you may have data that you want to extract from Salesforce Zendesk Google Analytics or other software as a service platform and The same is true for loading data to your databases and data warehouse you have to also maintain loaders for Postgres or Snowflake or Redshift Then there is the data transformation part Most API's return raw data That with tens or even more than a hundred attributes per entity That data is not properly formatted for running data analysis You have to find a way to clean the data Transform the data to a proper format and then define the dimension Symmetrics that we are going to use for your analysis And when most people think that they are done the really hard part starts orchestration Everything must blend well All the steps working flawlessly you have to be able to define pipelines for extracting loading and transforming all your data and Then you have to find a way to run those pipelines schedule pipelines monitor pipelines running Know when something broke or automatically restart pipelines that fail Meltano's vision is to take care of everything. I just described We apply the best practices of DevOps to the data management workflow People in the industry call that data ops So we provide an integrated workflow for data ops data engineering analytics and business intelligence and Meltano is a convention over configuration products as we are going to see during the rest of my presentation you can set up Meltano and Start extracting data from supported APIs in a few minutes just using your mouse That was what my tongue can do I forgot to change the slide so who are we building Meltano for I Described a lot of things that we want to do but we have to take it one step at a time So we are in the current version of Meltano. We're focusing on a single use case Our target persona the founder is a busy person at the startup using Meltano in single player mode As I'm going to do you make demo they have access to all the systems and data across the company But they are new to data. They do not write code or know how to write the square They need to do analysis to run the business But they are not a data analyst and they need to do both the data engineering and the data analyst Tasks because there is nobody else there to do that for them So in the current version of Meltano, we are focusing on running Meltano as a standalone application supporting well-defined APIs and Providing an end-to-end flow using Meltano UI our web-based API interface so that Someone can extract data from those APIs and load the data to a local post So instead of tell you what Meltano is. Let me show you how Meltano works I have installed Meltano have the latest version of Meltano in my laptop I'm going to initialize a new Meltano project Let's call it githlab hackathon Let's go to the directory and I'm going to start Meltano UI So the moment I get back This is our starting page here. I'm going to select an extractor I'm going to load data from githlab I'm going to use as an example loading data from githlab false So here I'm setting my private token and the group is githlab.org and the project in this case is githlab false Let's say that that's extract data for a month and I'm going to add a loader Meltano is downloading those extractors and loaders as plugins and installs them locally And as a loader I'm adding Postgres So I'm going to load the data to my local Postgres that I'm running Let's wait for Meltano to install the loader The moment I click save, the last part is the transformation. I'm going to run transformations Meltano comes rebounded with a couple of transformations And the moment I click save, a pipeline is ready for me with githlab Postgres and to run once I can run it on a schedule if I wanted And I click save and that's it. I went from having nothing to start the Meltano project and start extracting data from githlab false So you can see here the process running. I'm going to leave it running And I'm going to go to the last step of the process. You can see it here. We can return back later Which is analyzing my data. So if I go to model Models are Provide the starting point to explore and analyze data for specific use cases. You can think of them as templates that Provide only what's necessary for each use case. So in this case, let's say let's say that we want to check project starts for githlab false If I click here Let me remove this and that and let's say that I want to say check by milestone here and date And I want to check the total comments. So I want to check the total issue comments in the total Merge request comments If I click run Not start date. Sorry, due date Because milestones have a due date. I'm also going to filter my data. So I want To be a due date here, so it's not null and I'm not so going to Add another filter. I'm going to say that the due date is less than September 19 When we merged the two projects, so if I click here run and Increase that to 100 because we have more than 50 milestones And click on an area chart. I can see how Comments went for issues in the Mars in githlab false This is a very interesting report for me. So I can save a report so that I can use it later. I'm going to call it Comments per milestone Click save And then I'm going to add it to dashboard. Dashboards allow me to To have multiple reports at once and check them. They can be reports from different sources or so. I'm going to call this community dashboard Because this is githlab false and a lot of community members contribute to githlab false and Comments help me check that. I can also check other things like for example, if I go to merge a Mars per month I can check by merge gear month and Mars per offer for example, and if I click that I can see How many Yeah per month how many average merge request we have per offer which is also shows How the engagement from the community So I go I'm going to save this as a Mars per offer Per month, let's say And add it to my community dashboard and also create a new one because average and Mars are Interesting engineering KPIs But I have a second dashboard with my engineer KPIs. Let me show you one more Dashboard here report. Let's say let's use The bug report that I also did in githlab commit So I want to check bug reports and how we deal with them in githlab. So for example, I'm going to filter issues by And only keep the ones that have a bug label in them and I'm going to check by created Year and month The date An issue is created is when the bug was reported and let's check how many bugs we have reported and average days to close So if I click run and they freeze this 200 because we have more than 50 months here We can see here for example How the total bugs Went throughout the githlabs history and how we decreased the time to address those problems So this is also a very interesting report for me. Let's say that we call this bugs per month And we cannot this to our engineering KPIs that dashboard So now I have my dashboards. Let me show you Somebody like our founding persona can come here and check The dashboard so check the engineering KPIs and have everything together So the last part is that now that we have our dashboards and we can check the data is to have the data up to date All the time so I can go back to my pipelines And schedule you can see that the last job Finished and you can see here Meltdown extracting all the data from githlab Loading the data to a database and transforming the data So I can create a new pipeline meltdown remembers my settings So I don't have to add anything new and I can schedule it to run daily so that I have up to date Data every time I start my meltdown So more or less that's the end-to-end flow of meltdown and you can see how we try to help People to extract the data and analyze the data by not having to set up anything using a CLI or create speech or whatever So let me go back to my Presentation so what I presented was meltdown or running as a standalone up if you want to check it In the cloud we also support meltdown around the cloud and you can use the one click up Of meltdown or in digital or you can find one day one click up meltdown in digital ocean marketplace So it's very easy to set up a droplet and this meltdown of that way if you don't want to set it up locally Finally a quick glance at the underlying open source open source components That we have in meltdown or meltdown always built in Python And most of the components that we are using are also built in Python So for our extractors and loaders, we are using single IO taps and targets The senior eye community has done a great job of the great Reimplementing and maintain more than 50 Taps and targets and this is an interesting part that you can help us With contributing and adding more tabs that Mel Tano can support Mel Tano We are using for our transformation layer dbt dbt is an amazing tool that allows you to run transformations using pure SQL If you know you have not seen it please take it out. It's it takes care of everything for you So you just write your transformations with pure SQL and they take care of compilation Dynamic leaking testing and a lot more things We build at the moment ourselves the model and analyze part that you've seen Mel Tano UI But we also support other connectors like for example, you can use to Jupiter notebooks on top of our analytics database And finally for the orchestration part that you've seen with a great with Apache air flow that runs inside meltdown So how can you reach us Please join Mel Tano slack. I have a link. We have a link in Mel Tano.com. I don't post it here because it changes every month So go to Mel Tano.com and follow the link to join the slack Most all the team hung out on the general channel were very responsible Responsive please ping us there and you can also of course talk to us in issues in the Mel Tano project Or any other project in the Mel Tano group if you have something more specific to discuss with us How can someone help us First of all something that a lot of people forget and this is especially Important for us that we're at such an early stage. Please Use if you use Mel Tano and find it useful Please open issues for any bugs that you may find missing features or just ideas that you may have About what we can do in Mel Tano If you are good with Python, you can update one of our Extractors and not of the extractors that we currently maintain another missing API point endpoint For example in the github extractor are just presented. We don't export comments Notes as we hold them in github. So somebody could Go to the github extractor and not the API endpoint for getting comments from issues met request and epics and submit that Contribution if you are good with SQL You can play around we have ways to add more transformations and models The things I just presented if you build something interesting Please contribute back with transformations that maybe other people may find useful if you are You know some Python and after you understand how Mel Tano works You can help us by checking more extractors at it at the moment. We support more or less 10 extractors There are 50 more there We will get all the help we can get if somebody can check an extractor That means that you run the the extractor Check the configuration settings and go through the flow of adding an extractor to Mel Tano If you know of UJS you can contribute to Mel Tano UI and if you are feeling Good with Python you can contribute to Mel Tano code which is the core product of Mel Tano The core engine finally last but not least it's we always welcome any Contributions to our documentation and more tutorials for people to be able to onboard Mel Tano and Use Mel Tano in the optimal way If you're interested in contributing or doing a merge request for GitLab hackathon, we have a link here We have issues in Mel Tano group labeled as accepting merge request You can check them out. Let me open one So you can see here we have more than 20 accepting merge requests Issues you can jump into one some are Mel Tano core some are labeled as UI or testing or documentation So for example the one I was telling you is this one extra comments for issues, merge requests and epics you can get there And check the description and start working on a feature Or if you want to do something more UI based let's say for example, we have this which is pure UX It's a simple merge request Validate an extractor loader or something like that So let me go back to my presentation So who to ping we are a small team of six persons If you want to know anything about Mel Tano core or want to work on Mel Tano core It's being Mikael or me in Slack or inside GitLab Mel Tano UI is Derek and Ben For Vue.js stuff. It's also Ben and Derek and Ben is one of the most active Participants in the Vue.js community so he can give you a hint or two also For documentation me and Ben tutorials is me anything about extractors loaders transformations of models you can ping me freely And on what and how to contribute of course it's Dawe, Dawe is also one of the first employees of for GitLab So he knows well how to help people contribute and that guide people on that process And of course Danielle our general manager. She's very responsible On Slack and on our projects. You can ask her anything in here. Please contact us. We will be happy to help you out And I think yeah, that's it. Please visit Mel Tano.com and our projects check it out and get started Oh, thank you. Excellent. Thanks. Thanks for the talk and the demo And if people have any questions, I mean feel free to type them on the chat or or or verbalize him I mean the meantime I got a couple of questions for you, Janice, if you don't mind One is I mean so how so you're not Mel Tano doesn't follow the normal release cycle of rest of GitLab. Is that correct? Yeah, can you tell me about the release cycle and how it's done? Thank you very much for this question At the moment because we are at such an early stage with the rate very very fast. So we have one milestone per week So with targets and with the rate weekly and that means that shows how fast with the rate and also that there are Small things for everyone to get in and help us out or do stuff Cool. Cool. All right. Excellent. All right, so anybody else have any questions or I mean, Derek, anything else you want to add Nothing much to add. But yeah, definitely. I think down is trying to reinforce that we had just ping us on slacker in issues I don't want to hear from you Yeah, and otherwise, you know, it's a great job explaining basically everything you'll need to know but one of the big differences between the experience between Mal Tano and GitLab is that there is so much low hanging fruit in Mal Tano. So many things were just five minutes to work or half an hour could get us so far Well, of course, GitLab with the skill of it and just the fact that there's like at this point hundreds and hundreds of people working on this big time It might be harder to find that little thing, but if you just want to kind of work on your future skills or your private skills, you just want to try a new project where you as a contributor would actually If you become a Multana contributor, you immediately are one eighth of the people developing Mal Tano. GitLab is not the case anymore. So in that sense, it's a really great way to kind of become part of Get a contribution community of a new project and there's so much to do and so much to learn and we are of course more than happy to help anyone out Figure it out with us basically I appreciate it, Yanis basically listing everybody on the team on the last slide. So you can you can ping everybody in one shot, I think I mean, the other question Yanis was So I'm sure this is documented in the documentation. So what are like the system requirements like for for having Mel Tano on your on your system on your laptop. Oh, they are very low So because the iPhone and You can get started very easily. It's the requirements very low. The only part that requires something to be done is the Only part that requires some memories running Airflow and running a Postgres case, but you can do that even with a small droplet if because I would mention Cool. Excellent. So yeah, so it should be relatively easy for people to get started. Cool. Yeah, it works on my 2010 Linux using Google to 16 has no problems. Cool. Nice. Cool. Yeah, I don't see anything on the chat. Any questions or comments from wider community members. Maybe I can say something. I definitely want to be part of the one eighth of the contributors. I noticed that there were a couple of good club extractors. I was wondering whether Mel Tano started because of of a pinpoint that the club had internally. Yanis. So you want to answer that. Oh, feel free to. I mean, I think you have more context here, but Brian, I think it's a good question. Yeah, I mean, get let get like wouldn't be, I guess, in a way investing in this project if it didn't make sense at some point to also get in the business of solving this problem and that very much happened when Get Lab itself found that there wasn't a tool out there that did what get labels looking for. And then of course it's easy at certain points say, then why don't we build it with the community. But I know that Yanis actually, I mean, I only joined the Mel Tano team at this point two and a half months ago or so. Yanis has far more background information on how it got started and how he got involved with it in the early days, way ahead of me or Danielle or any kind of, you know, kind of more start the hierarchy system coming in place. Yanis has been there forever. Yeah, so we started Mel Tano with our general manager for a year was seed. So he was very involved. This is this was his idea, his vision of building a product and bring the devil's mentality and what we are doing in Get Lab to the to the data management world, because if you check how data management works at the moment, especially the part extract loading, transforming and all the stuff that are very complicated and they require a data engineer to run them and so we wanted to bring all the mentality of Get Lab to data management. And that's the core vision, the core idea. And that's where we are focused on and we are keep our vision still with Danielle and the rest of the team at the moment. And I would say it's not a coincidence that our persona is called the founder, because in this case the founder literally does kind of basically reference or for said, maybe not the set of today because of course the founder of we're focusing on a far earlier stage kind of people of the company know data team know nothing, but access to data sources and wanting to do, you know, something with it gets some dashboards get some insights out of it. So we are well sit one of this because he was a founder and he couldn't find it and now we're building it for founders just like so. Oh, awesome. I think I final closing, as I was saying, there are a lot of issues on Get Lab, but there's a lot of demand for them. So I'll definitely take on that. Awesome. Can't wait to see you on site, Brian. Yeah. And you too, Matthias is also in the, in the zoom. Cool. Yeah, Brian, I'll follow up with you on it. We have a call schedule a couple of weeks. So, I'll ask you how things are going. Cool. All right, if there are no other questions. Thanks everybody for your time and have a good rest of your day or evening. Have a good one. Cheers. Thank you.