Live Stream of the in person meetup.
https://www.meetup.com/pydata-suedwest/events/296355526/
This event will be in English.
Talks:
1. Transform Your Business: A Roadmap to Being Data-Driven with Data Ops – Simon Pressler (Königsweg)
2. Data Versioning with LakeFS and LakeFS-Spec – Max Mynter & Jan Willem Kleinrouweler
3. My Data Stays Here! Implementing AI Assistants in a Data-Saving Way – Hans-Peter Zorn
Agenda
18:00 Doors open
18:30 Welcome – Alexander CS Hendorf
18:45 Transform Your Business: A Roadmap to Being Data-Driven with Data Ops – Simon Pressler
19:15 Data Versioning with LakeFS and LakeFS-Spec – Max Mynter & Jan Willem Kleinrouweler
19:45 ️ Break: Networking with snacks and beverages
20:30 My Data Stays Here! Implementing AI Assistants in a Data-Saving Way – Hans-Peter Zorn
21:00 End
—-
Talk #1
Transform Your Business: A Roadmap to Being Data-Driven with Data Ops”
Simon Pressler (Königsweg AI)
Many organizations ask themselves: do we have a data treasure? – how can we find out and how do we start? In the current data-centric business environment, this question becomes increasingly relevant. Drawing from extensive experience in the field, this talk addresses those who are enthusiastic about data yet may find themselves at the outset of their journey, or uncertain about navigating their path to success. DataOps, an interdisciplinary field seamlessly integrates data analysis, software development, and system operations. This session aims to provide clarity and direction for unlocking the potential of your data assets, marking the beginning of a transformative expedition into effective data utilization.
Simon is a Data Scientist holds a Master’s Degree in Comparative and International Studies from ETH-Zürich and a Master’s in Data Science from the University of Mannheim. Outside of his professional life, Simon is passionate about long-distance hiking, a pursuit that showcases his dedication and resilience
Talk #2 ️
Data Versioning with LakeFS and LakeFS-Spec
Max Mynter & Jan Willem Kleinrouweler (appliedAI Institute for Europe)
Data versioning, essential for reproducible Machine Learning, involves maintaining and accessing various versions of data sets over time. This talk highlights LakeFS, a tool enabling Git-like versioning in data lakes, and LakeFS-Spec, a Python library developed by appliedAI Institute for Europe. LakeFS-Spec, compatible with FSSpec-supported packages like Pandas, allows intuitive LakeFS interactions and offers advanced features like client-side caching and direct file system operations.
Max is an MLOps engineer at appliedAI Institute for Europe. He has a background in Physics and Social Sciences, spent some time as a visiting scholar at UC Berkeley’s School of Information, and had previous stints as a Data Scientist at an energy-tech start-up and at Allianz Global Investors as a Quantitative Risk Analyst.
Jan Willem is the Head of ML Engineering at appliedAI Institute for Europe. He received his PhD in Computer Science from the VU University Amsterdam. Before joining appliedAI, he worked as applied researcher and portfolio manager in the fields of media delivery, mobile networks, edge computing, and IoT.
Talk#3 ️
My Data Stays Here! Implementing AI Assistants in a Data-Saving way
Hans-Peter Zorn (inovex)
ChatGPT has become an integral part of many people’s professional lives. Nevertheless, many companies are reluctant to provide their employees with enterprise chat solutions. In this talk, I will demonstrate various ways in which such a service can be implemented in a data-saving and data protection-compliant manner. Be it through a private endpoint in the Azure Cloud or open source models under your own control.
Hans-Peter works as Head of AI and CTO at inovex to help customers overcome their challenges – sometimes with the help of AI. He studied computational linguistics in Heidelberg and computer science at KIT. He then spent a long time working on speech dialog systems, natural language processing and big data architectures.
A big thank you to our sponsors:
appliedAI Institute for Europe, for hosting the meetup.
KÖNIGSWEG, for supporting the organization.
Contact
If you have any questions or suggestions, please feel free to contact us via:
Meetup
Want to speak? Submit a talk here. http://bit.ly/43QBJkn
Interested in hosting an event? Here’s our Info-Deck & contact to the organisers! https://bit.ly/pydata-sw-infodeck
So okay so welcome everyone to uh the first P data in h one so P zest this the first time in h one so welcome everyone um and some of you know already the procedure some of you might not um so this a disclaimer we recording the
Meetup I we streaming live on the pi data Channel on YouTube or LinkedIn on Twitter also we recording also like audience questions so yeah so you have been informed um yes so this is the 28th pataa zwest pataa zwest started uh as a collaborative meet up from yeah actually
After the Cal conference we said okay pataa now is with the Berlin Conference let’s do something local and so we had Caro idle at Manheim and now we’re really happy to include hyron as well so and um so welcome everybody here uh at home and at the screens and this brings
Us to the people watching online and the stream uh so hbr what is hbr hbr is in Southwest Germany as you might guess from the Meetup name and some of you already know this slide where tell hey where are we located in Germany and I let me show you like the the
Distribution of universities and the best universities in Germany and you see there’s um it’s not everything is in Berlin because if I’m I’m inational conferences people always ask me oh you’re from Germany you’re from Berlin no I’m not from Berlin so I’m from Southwest and we are here in highw one
Which has also a university and it’s actually a collaboration with the um university of Munich that’s why there’s this little uh line there um also this is the distribution of companies the most most important German companies you see we have like um yeah this is the distribution speaks for itself so no the
Economic power is not in Berlin we see we have all yes yeah some people don’t know that especially International so we see we have all the German um Excellence um companies down here in the South many people many companies you know like Bosch or like porer and and so on and so
On um and yeah we’re just like basically in the middle of it in hardland of the German industry and this brings us to our host eai and would like oh sorry oh no I think uh no sorry sorry that’s that’s I added that slide but you
Will be there Min uh let let’s go to that so um I think it’s very important this is a very inclusive and open community so please raise your hands if you were here around at the first time at the Pata Meetup wow very warm welcome to you welcome to P datas Z West
Um and so everybody’s a name tag yeah Nikki took care of that um it’s very important this is an open Community there are no stupid questions just interact with people talk there’s no stupid questions at all like the best tricks I learned from chatting with people at pataa
Meetups in the breaks um and we have this Pacman rule so what’s the Pacman rule the Pacman rule is people stand in circles and always leave a spot open so somebody else can just join in and also please invite them hey join the conversation let’s talk and then you meet new people
And learn new things so pataa zest um as I already mentioned pataa is uh was initiated by n Focus num focus is a us-based foundation which supports many um scientific packages uh you probably might use a few of them or many of them in your daily
Work or research um uh so um no focus is all about um scientific Computing and supporting open source um we have a code of conduct uh so as I told you like we don’t we do not tolerate any harassment in the of community in any way uh if you
Have questions there’s the non Focus code of contact online if you feel uh if you think there’s something you want to report or if you think something inappropriate happen just talk to the organizers like me or Kevin on video Ops here behind the scenes yeah and or you can also like report
This online so it’s very important everybody feels welcome and we yeah so uh this is the Meetup we are just like hi here in the middle um in pataa zwest uh we also have a YouTube channel we’re streaming to the pataa YouTube channel um it’s very important
For us to record conference talks also if possible the meetups and share the knowledge on on um on YouTube and this is not just like hey YouTube’s cool so actually I was sitting in the train to berin to the pon the in P data conference and I was chatting with a
Colleague and we mentioned some stuff and there was sitting somebody else just across of us and said oh yeah why are you talking about this and I said yeah and he said yeah I’m from Nigeria I watched your videos and I learn from them and so this is really working it’s
Not a theory it was really like yeah the best proof ever um so um hi West is already 5 years old crazy isn’t it we are almost 1,00 members now um so um ping us if you want to give a talk um we always look for speakers um today we
Don’t have lightning talks because we have three talks uh but if you want to contribute a talk or lightning talk there’s a link you can also check the QR code or it’s also on the Meetup site and of course if you also work for a company
And want to host the meet just ping us because we’re also look always looking for hosts and this brings us to today’s sponsors um we have uh applied AI uh providing us the food and drinks with have the Epi for providing the venue and Kix week for um supporting the
Organization oh no food today that was sorry that copy paste that was from the Big Data barbecue sorry for that so um so please welcome ahed from EPI and he will tell you a bit more about eepi thank you you go so hello and welcome we are happy to
Host you for the first event in hbr in Epi space and I will give you a little introduction about what’s EPI and what we are doing so first I’m Ahmed I’m part of Epi I work in the corporate real estate uh Department I am responsible for the building information modeling
And all the topics related to digitalization especially digital twin I know digital twin is kind of buzz word nowadays but yeah because we are developing a digital twin for the main campus so who are we we are trying our vision to be the global home of human AI
That’s EPI and we are developing eepi gradually so we are now in the first phase of Epi which is this building so that’s the phases uh step one or first phase that’s this location with 1,200 M Square next phase I’m not sure if you saw it but it’s just the building across
Here next to this building it’s called velf so that’s the second phase which will have 6,000 square meters and it will be open in 6 months this year and then the main vision and the big vision is the eepi compass which is the first phase of it which is the part in the
Middle should be open in 2027 so here uh we are are building the initial ecosystem with the first strategic partner and that’s part of the our uh members and partners here and second phase velf that’s the design of it and yeah that’s all the amenities for it like the spaces so it
Will have real lab uh also an app space which is very important for everyone there and yeah we are also it’s for us also the test phase for the main campus so we are trying a lot of new ideas related to digitalization and also for uh the office uh in inside offices inside
Also and then the third phase which is a big uh AI campus and uh with the help of all our partners members we are also uh being with the uh stateof art Technologies we are trying to make it like a home of AI in all Europe so thank you and I’m happy to
Have you here and also to be a participant in the meeting with you and thank [Applause] you I’m leaving this yeah no problem so yeah thanks also from my side for hosting uh it’s really great it’s also like we have plaman today no pizza and
Now I would like to introduce it to Lan Yan Willam from a FL [Applause] AI see if I can clip this on somewhere hi all welcome you all and it’s very nice to see you um with P data here uh here in hon my name is Klein I’m the head of
Ml engineering at apply and I will say a few words about basically who we are and um and what we’re doing so PDI is a nonprofit organization we have slightly more than 30 people and our goal is to create and share knowledge on how you can apply the latest AI technologies
That are around there and specifically we’re focusing on doing this in a reliable and in a trustworthy manner so what does this mean this means that for instance we create insights on the AI act and see what can this do for you as a professional and how can you apply
Thei in a in trustworthy fashion we also develop open source software to make it easier for you to do this we have training programs and courses this is for students or for uh for young professionals and companies around really to do basically upscaling um of
Uh of AI we um have sort of AI Innovation insights showing what is happening in the startup landscape um Etc and we organize events on bringing people in the ecosystem and in the community together we organize seminars we organize U events and and meetups and also Yeah by sponsoring events like um like
Today now if you are interested sort of in what AI is doing and say hey I would like to also attend one of those events you can uh find more information on basically any uh social media platform of uh of your choice and I just want to
Wish you a lot of fun today learn a lot meet a lot of uh a lot of new people and just have a good time yeah thank you very much yeah so that brings us to Z from Kix week here you go thank you thank you
Very much hi I’m sir um I’m CEO of kwick and together with Alex I’m here and super happy finally to be here in Halon because uh I personally uh family background is not only in heidleberg where I located but also in Bavaria so I’m many many years I now go back and
Forth on the A6 and I always thought like well okay what about Halon and now finally we are here and I think it’s the thrill thing after 5 years that we together founded P zest so coming from Que phix week we are a data n AI um
Consultancy company as you can see here we are super passionate about uh everything in this field and that’s why we uh in addition to that also support from the uh very logic of our internal DNA anything around open source and that’s why we contribute into Pata with
West but also into uh Pyon um and anything about Community um so whenever and wherever you might be interested in any kind of collaboration with please uh get to our homepage get in interaction with us uh discuss with us and meet us in person during the breaks have a nice
Drink together um I will now not talk too much about what we do in concrete projects or whatever why I’m super glad and happy that Simon of our team is today here and he will give us a little bit Enlightenment of what’s going on on a day-to-day operational basis
Nevertheless I wish you all a very very fun evening today here and looking forward to commuting it and chat with you later thank you very much thank you to you so we also have a LinkedIn group um this is moderated so no recruiters no spam um it’s just for us so please join
Um this is the agenda we hear the welcome we will soon start with the first talk uh we have two talks in a row and uh then we have a break and the break is in the third floor and the good thing is we don’t have pizza to today we
Have oh what is it FL yeah which is FL out in English which is more like French actually to me um yeah so uh yeah um without further Ado we come to community announcements now so who knows this building you know oh yeah only two oh
Three four yeah okay this is uh this is the inside of the building and here you are see on stage an a colleague from uh from y Yan villm is Paula um she also unfortunately she’s sick so get well soon Paula if you’re watching um we’re here on stage it’s a Community Driven
Conference at the BCC in bin and good news ah it’s coming back in April and uh actually we had a record submissions of proposals almost 500 which only says like only 20% will be or are accepted so um the program will be released very soon um tickets sell fast so better get
Your ticket um now than never and we have a call for volunteers on site open which is can be found under this link or you can just like go there and subscribe to our newsletter to be informed about updates and this year we also have a really great celebration it’s actually
10 years of Pi data bin which is pretty cool we’re doing this for 10 years now um there’s also pady suit vest locally um there’s going to be um a meetup on March 6th for the international women’s day um just go Um on Meetup and look for
Pades at West joins it’s um pades is um enabling women in Tech and we also I mentioned we always have a open call for speakers for meetup so share your experience with the community five minute lightning talk a full talk yeah um the Meetup is safe space you can
Uh train for a big conference talk or submission later and uh organizers are be also happy to mentor and help you um with the talk and I think that brings us to my colleague Simon which will talk about how to transform your business a road map to being data driven with data
Ops and we know switching it’s all going to be F and yes all companies with clean data so yeah oh thank you very much and indeed it worked nice so also welcome from my side I’m the last part in the trio of K who is here today and I’m going to speak about
A road map to being data driven with data operations today and I will do so in three steps we’re g to start with something introductory what is data Ops why is data Ops really cool why do you want to have data Ops then we’re going to have a
Look at the second part how to get started how to identify potential in your company and then we’re going to spend half of the talk on something practical where we’re going to have a look at two use cases so you only need to make it through the first 10 minutes
And then we’ll have something practical so without further Ado and then dragging out the theory stuff data operations 101 and for this talk I wanted to do everything very nice and clean and I googled around what’s the textbook definition of data operations and as typical in AI it seems
That there are 100 definitions if you ask 50 people so I think it’s most instru Ive actually if you think about data operations as the data machine room of your company you can think of this analog to let’s say a large cruise ship you have a machine room that provides
Power and propulsion to you but you don’t want to have to run down there if you’re the captain off the boat and get your hands all dirty every time you need to change course so you want to have a machine room that takes care of all the technical stuff and therefore the goal
Of data Ops is have a dedicated unit ensuring quality and availability of your data you want to enable people to work with data without bothering them too much with all of the data management hassles and to stay in the analogy with the cruise ship not every cruise ship is
The same depending on if you’re cruising in the Antarctic or if you’re cruising in the tropics if you have 100 or 6,000 passages on board data operations as well needs to be tailed to your specific needs I’ve given you the two current hype examples a data mesh and data Fab and very abstract
Here um with this out of the way why is data Ops really great it’s great not only because it improves the availability and reliability of the data for the Departments working with the data in your organization but also because if you implement data Ops using open source software you can establish a vendor
Independent way to deal with your data so you’re saving potentially a lot of money um by avoiding licensing fees if you’re going open source and you can avoid vendor lock in which can be really handy in the long run um you can work at your own pace prioritize what you need
Exactly when you need it and you decide which data goes where which I guess will be the talk the theme of the third talk um so you might be wondering now do I need data Ops and here I brought you something that we’re using internally on
The bottom left side you can see how we as kisg are categorizing different companies and it ranges from data naive to being highly AI optimized depending on the level of technological Improvement in your organization and the level of how important data is to your core business processes and I’ve brought
You the six core things here for being a data driven company and you can see you’re relying heavily on Data Insights for decision making and data analysis and AI are becoming Central to Core Business processes so you got to make sure sure that these things work and that’s where data operations is
Handy um the only problem with all of this is this is how we come up with the assessment the whole thing is not exactly straightforward and that’s why we’re going to have a look at how to get started in your specific case in the second part of this
Talk uh which I labeled your data treasure map because in order to get started with data operations you need to identify two different set of things the first thing you need to identify is the treasure itself how much data do you have is your data AI relevant what types
Of data is prevalent in your organization how many data points do you have have you a history with your data like is it historized all of these things and you might be saying now of course I know easy no problem and I would contend that if you’re working in
An organization ation which is just now moving into being data driven and is in business for more than five years probably you don’t know that’s the funny thing um usually there is not only significant heterogenity across departments in data quality and quantity so if I ask you working in Department x
What does Department y have actually in terms of data you might be surprised but also there tend to be a lot of CEOs and what I labeled knowledge hoarders these tend to be very dedicated people taking care of their silos guarding it enviously and as soon as they retire you
Realize that they were the only person capable of handling the data in this department specifically which brings us neatly to the second thing you need to know which I labeled your data Black Market that’s another thing that’s just bound to happen if you don’t have good data operations in your company and you
Have been in business for a while you know things just happen put yourself in the shoes of somebody who needs to make a report you’re asking your colleague from sales hey can you give me the newest sales data he or she sends over an Excel file you’re making the report somebody else
In sales didn’t add their recent sales to the list you go into the meeting the report is already outdated two days after it’s been made nobody can trace the error nobody knows what’s going on so you see these unregulated d uh data flow are badly traceable super error prone produce silos redundancies and
Breaks super easy if you colleague from Mark from sales goes on holidays do you just stop making reports or do you ask somebody else who’s then digging into a colleagues emails it’s it’s all messy but at the same time these undocumented data flows tell you something very
Important they tell you how data in your organization would need to flow an ideal scenario they show you a need so now you might be wondering that’s all nice and well but how do I figure this out and the unfortunate truth of it is shoe leather there is no easy way to
Getting all of this stuff sorted out beside talking to people which can be quite uncomfortable at times that’s the second point on the slide you need a neutral view to get an idea who is guarding which Silo and depending on how um what’s the polite way to put this depending on how
Passionate they guarding their Silo this can become quite unpleasant but if you are going through all of this hassle what you are left with in the end is a great starting point which basically tells you where you can create a lot of value with relatively easy data operations
Measures um so you got to prioritize and choose accessible and highly valuable projects to your organization and then make sure you’re not creating lighthouses keep in mind it’s never about just creating a notebook which will then be archived away once the presentation is held try to create infrastructure and this has been quite
Abstract so far so let’s uh walk through two projects we’ve been having recently and we’re going to start with example one where we are we’re developing an AI application to improve planning processes we were having a customer um and they were in the space of planning large multi-year
Projects and the problem is that Staffing these long running projects is relatively complex if a project runs for more than a decade you got to hire people you got to make sure you have the right people for each project phase and getting it wrong can delay these projects and cost Millions so
There is a real need for to solve this issue at the same time what they were doing is every project planner took care of each project individually and came up with numbers controlling said sounds about right probably or was asking questions back but it was all based on individual
Experience so we set out to provide an application uh providing a reliable Baseline estimate so we didn’t want to automate this process but rather leave enough space for planners to adjust it up and down but take away the work of doing it all from scratch every time and
Just leave them with an up and down adjustment in the end so just remove the redundancies but still allow for individual expertise so we started working with the customer and this approach has the key advantages if you’re going for a high value application in the beginning that you’re creating immediate
Value this can help you to get skeptical actors on board because you’re immediately showing okay if we’re transitioning away from silos we’re actually creating something that will benefit everyone in the organization not only in the very long run but already in the medium term um also if you get started with a
Specific application in mind instead of just doing broad architectural thinking you will run into issues that you didn’t see coming which is very valuable in its own right because it tells you even more about your organization’s data that’s what I put down here on the slide as the last point point it can
Serve a cryst as a crystallization point for future improvements an example we got started with our application in this instance we achieved good first results and we thought hey just add more data let’s improve this let’s make this really great however we soon ran into a couple of data problems including an encrypted
File without exess key the only person who ever knew that access key recently retired which was a little bit unfortunate so we had to figure out a way around this we worked with an orphan data warehouse so we could only take data out of it but nobody knew how to add data into
It and important data sources were not integrated and especially stuff like the Excel files you only really ever notice if you start building an application just from theoretical planning you will not get there talking about integration this brings us to the second example and this example highlights that
You need to tailor what you have what you want to do in terms of data operations specifically to your use case in this case we had a customer with strong data silos plenty of expertise in each department everybody’s super competent just the communication between the different department in terms of data was severely
Lacking and we tried to solve this problem by implementing a data mesh to keep the data in in the hands of the local experts but at the same time create reliable access throughout the organization so we eliminate the need for informal data flows and if we replace informal data
Flows we need formal data flows so we set up a bunch of ETL pipelines to automate the data movement from the data within the department to the provided API outwards facing um we implemented some data quality management measures like automatic data validation and contractually defined standards and these helped us to create
Trust between the different departments so you don’t have to check every time you’re in contact with your other department did they give us the right thing does it have the right format no you can just develop you know what you get from the API and this alone already alleviates a lot of inefficiencies
We did all of this dockerized deployed on Ms Azure uh for the obvious reason we wanted to have great scaling in the future uh we deployed terraform to keep the cloud configuration reproducibly versioned just to make sure that we know which um software version goes with which infrastructure
Version and doing all of this we were still hyperscaler independent which is really cool because we implemented it all using open source software not relying on proprietary Services by the cloud uh vendor so if our customer ever says let’s move this from Azure to Google I’m not saying it’s not a lot of
Work but it can be done which is really cool I think that’s it from my part in case you’re having any questions I’m here do we have time for questions what’s have time for [Applause] questions we have question yes but this is only because you should really think about who you L
Your stuff so thanks Fabian for never returning the thing here but I the the professional thing here but we can still improvise with this cooking stick to microphone Alex you look threatening yes I’m very sorry and thanks again Fabia ban if you’re watching yeah uh yeah questions no questions ah
Question yeah first of all thanks a lot for the for the talk very uh very nice um I was just wondering about um sort of the the data that gets sort of stuck in Excel files I mean you hear it more often and it’s there because people like
Excel how could you convince them to turn this or get the data out of it and turn this into a data mhine ETL pipelines and um all that that’s the the problem if you will it’s not only about creating infrastructure but you also have to change data
Culture um you got to get people to understand that having several copies of a Excel files and working off of them instead of using the API to draw the data a new every time creates a lot of problems and it just takes significant amount of convincing and just general upskilling
Making people more data aware that’s part of the that’s the non-technical part of becoming a data driven organization that’s the culture aspect if you will oh actually if I might add to that we ask them hey can you please look in all your papers or somebody noted put down the password because the
People were really in retirement so yes so that that’s pretty convincing as well like it’s good to have have a bad example at hand yeah it’s a bad practice so um yeah more questions yes another a question thank you very much for your talk I was wondering so
Since there are so many Excel lists going around and um which techniques do you use to turn Excel files into reliable databases um that depends on the Excel file to be honest if it’s something that has been passed around multiple times already in the organization is probably best just
Work off of the single point of Truth just get rid of it and the other scenarios you just got to feed the Excel file into your um into your database at one point I think that’s the only reasonable way to go edit to the overall pool of
Data transition away from using Excel to store important information generally so it’s basically just there are already all of the libraries and it’s just an API call convert Excel to SQL or how can I Alex you want I may add to that we we like to put the Excel files into pet and
Work with pet files and eventually also integrate pet files into a data Hub so make them more accessible but that’s also just like an intermediary solution because yeah we don’t really want to work on XX file or single file formats except we say okay it’s a big solution historical data that’s something
Something else you and actually like integrate it into um like a database data hub or like an yeah an organized and data managed system so sorry for the slid question uh yes just you got you got an image doesn’t here oh maybe I um that’s unfortunate that’s unfortunate but okay in the
Meantime maybe I can ask the so how do you manage to be vendor independ and at the same time to provide some authentication and authorization system for personalized users for applications um for Alex do you want to take this question sorry I didn’t get the question
I was uh a question sorry so how do you manage to be vendor independent that you don’t use the services from Asher or UCP and at the same time you implement authentication and authorization in a proper way I mean we we use open source for everything so of course you will never
It’s not just like okay you cannot use the same terraform of course to have multi multiple hyperscalers so there’s a certain vendor I wouldn’t call it login but like inflexibility but you can fix that because like program code dockerized um we like to work with files
So um I think uh when the logging in the cloud happens if you of course use their databases use their extra services this is where we Define window login so of course like with author authentication and other services so it’s it’s it’s a bit more of course moving from One Cloud
To the other is still work it’s not just like hey switch unfortunately all right then hope the next Works say next talk is Max but yeah please give it yeah sorry for that yeah Max has a microphone and we will see what happens to the blue screen
Here now that we at such a smooth transition in one way I hope that we can reproduce that yes no I think yes let me check my my resolution is reacting to something yeah so thank you very much yes yeah so please welcome Max from applied
AI uh please stay here with the camera like this here okay away and yeah welome Max voice talk oh it’s dat oh okay that’s bad so that work for for uh just let me try something because sometimes it’s just like sorry for the break yep SP yeah in a
Bar Banger the crowd crowd [Applause] I’m yes HDMI is one of the best standards ever you can see near