25th OpenAIRE Graph Community Call

So hi, good morning everybody. Welcome uh today to the 25th open air graph community call and thank you for joining us. Today we have with us Andrew Niki Pavo who is industry leaison at open air along with other representatives from open air and alienai uh who is going to present the winners of the open air AI hackathon. Sorry one second just admitting some more people coming in. Um great. So we kindly ask you as always to please keep your microphones muted during the presentation to avoid any background noise. Um and we'll have time at the end and also maybe a bit throughout this session might be a bit more informal as we're discussing you know winners with different presentations uh for questions. So if you have any questions you can write them either here in the Zoom Q&A or chat or in the Google doc which you'll find in the chat which I'll paste in there right after. Um, and then every any questions we don't get to today, we'll answer directly in the Google doc. And a reminder that today's session will be recorded. So if you do not wish to be recorded, please let us know and we can edit any questions or comments you may pose out of the recording. So with that said, let me put the Google doc in the chat and I'll pass the floor over to Andrew Nikki. >> Yes, thank you very much again. So welcome everyone. Uh as you know we organized the hackathon some few months ago two two almost months ago and the deadline closed 20th of August. We had really good ideas. We are we were really happy to review and see you know what you had in mind how you implemented it. And today is the day that you know we we we announce the winners per theme. The grand prize winner and the community vote winner. um my colleagues may be a little bit late today because they have also many calls at the same time. So you know we will be patient and also I know that some of the applicants also follow a seminar so maybe we will ask them to unmute for a few seconds if needed. Um, as you know, we run this open air graph community calls for a long time and the hackathon pointed on the MCP AI uh service tool uh from Alien Intelligence because this is a collaboration now with alien uh that has lots of experience on how to use properly AI uh you know take all the information and use it in the right way ethical, legal And we also use the open air graph version 3 API uh which is the latest for this hackathon. And uh we wanted to test also how users comprehend this new uh functionality now where you can go ahead and write prompts somehow connect with the graph in a new interactive way as we do it in all these AI tools you know already in the market. This is something new for us and uh we wanted also to see if if we find more value through the open scholar works and also if we find something that we haven't considered hidden value or maybe a bug that we can fix in the process you know through the graph and through the through the pipeline of how the graph aggregates and then offers all the information. So thank you again for this. Uh I won't keep you waiting for a long time. So uh I will start I hope you can see my screen uh with the logistics. So we we received many uh people uh you know uh registering to the hackathon at at the beginning but the valid ones uh which means people that did actually uh you know went through the process correctly there were 18. We had people from 11 countries which is very good to know which shows also how powerful can be you know when you have an open hackathon open data open science using open-source information uh 14 affiliations which means you know people from universities research centers uh solo innovators developers uh some startups uh we had 12 uh submissions uh when we closed the hackathon and why I mentioned this because we also received one more application from uh a user that didn't pay attention to the dates and we had a submission after the closure of the hackathon. uh but we asked that person you know uh that we cannot take it into consideration for the hackathon but of course we can discuss it uh later and see uh the value of the application and we we went through the evaluation process uh and we have nine finalists and why we have nine out of 12 because we had a threshold uh people uh collecting more than 16 out of 30 uh uh score. So we reviewed the 12 uh submissions. We were five people uh three from open air and two from uh alien. two from Megan were Gen Deabi and Leo Bordell from he's the first one is the AI head of AI solutions and the second one is the chief science and technology officer of alien and from open air we had Paulo Mike the CTO we had Clauddio Jory who is the service manager and engineer of open air graph and then it was me uh that become the industry leazison. So we we tried to to merge uh our our scores and to view it from the techn to view everything from the technical side and also from the strategic side. So I I don't have a good powerful you know technical background but I I was looking more on the value the impact if the user can understand it better. uh and the technical team was working more on uh the code, the documentation, uh the data, this kind of stuff, the usage of MCP and AI uh and API. And as I said, we have the nine finalists. H you received an email yesterday, the ones that have this uh score up to 16. Uh [clears throat] sorry. And then as I said here are the people that had uh this review. So we criteria one four was pure technical. Five and six had strategic and technical and two and three only strategic. So here are the nine finalists as you know by email yesterday and we will circulate this again on social media. Uh number one was the open air evidence ledger. Number two, open air graph evaluator. Number four, from mathematical reasoning to algorithmic judgment. Number five is the replication rather. Number seven is data footprints. Number eight is the open quantum evidence adgas. Number 10 Sophia in the graph. 11 open gap. And 12 reproducibility bands. Now the community choice. I'm going to reveal it. I did this fast. So we as we asked people to vote. We had uh a um we had a script running. So we wanted to make sure that only one person votes. So the community chose winner is number four is from mathematical reasoning to algorithmic judgment. Uh I am not sure if Elen is here because I I don't see now the list of uh people here. If Elen is here, of course, we will email her. uh C is the one who got you know more more people voting on this idea and as you know um we had on the website we had uh many uh boxes small boxes for each application and links to the demo the use case the code everything that was critical so anyone can have a look and and check it out. So, congrats to uh to this uh applicant. And I think I should go to the next one which is one winner per theme. So, theme A the winner is number four again. theme uh B the build which which had I think the the mo most of the people who applied for theme B uh theme B is the replication rather it's the winner the replication rather theme C the winner is open quantum evidence applas and theme D the wild card it's the repability bench so congrats to all of you because we had really good competition mostly on theme B. Uh it was really intense. Some some of you had close scores and we had an extra discussion uh on Monday to see you know uh that we all align and we will be you know fair with everyone and we make the right decision. So thank you very much. Of course, as I said, we will post more on social media now, one by one uh what the winners did, you know, uh the information they had shared. Oh jeez. And now uh I think I I should proceed with the grand prize winner. Um and this number and it's here. I saw I saw it when I entered the call. Data footprints. So data footprints it was a team with Ana Kayas Ignasi Yastida and Juan Jose Baut I hope I pronounced it correctly. Uh it was theme B theme C apologies theme C analyze it had a score around if I remember correctly 27 something out of 30. Uh so what the team wins uh €500 uh they will have a pilot partnership status you know we will co-develop with alien team of course and uh they will be of course promoted and featured uh at the next open air big event open science fair uh with travel grants and uh we will have a followup to see if we can really promote this with any of the funding opportunities uh that we have at open air through projects through collaborations uh something that might came up in the next months. So we we will have a good discussion on this. Congratulations to Anna and the team. Uh it was really really very very nice uh what you created and why you know few words why we had this selection uh it because it was you know it was pointing on a very important uh topic on the re research assessment you know we have indicators we have uh citations we have the beep indicator on the graph and uh it was something really new to us to see you know something on this topic because it's really important and also we saw that you you used MCP connector the graph again API and it was a good uh architecture uh also uh delivered and also we we we saw that you use the orchit ids uh through the pipeline as as it's it says here because you know many of uh all the authors that have these orchid ideas, they have something that follows them with ideas. So it was really important to follow the assessment through each person again and be able to to see and verify uh who is who is who created something uh where that goes uh where it is connected and then go back again. So you could go from one one direction to the other and reuse it again and go back for go back and forward actually. So thank you so much. It was really great and everyone agreed you know all all five uh evaluators agreed on this and so what you receive just to remind you know the theme winners number four 5 8 and 12 uh they will get 12 months uh of alien gateway credit. So the demo that you used is uh when you logged in when you signed up to the hackathon also has uh a dashboard if you remember. So what you can do you can use it for 12 months that gives us plenty of opportunities now to improve the tools to to work them out. So you have time to do something you know to improve uh reinovate recreate develop it better. Of course, we will be uh close to you to help you out and um also uh we will see uh what we will need in that case. Uh you will have a community graph uh call focused on on you which means that the future community calls of the graph uh can of course show in in more detail uh what you created. So our viewers and online and also the people that participate can can have this chance to have you know questions. Uh you get feedback and uh you get promoted through uh open air. Uh the community choice vote number four uh get six months of the alien gateway credits and we will have of course as the theme winners. We have uh communication channels of alien and open egg you know promoting what you did. We can we can write some blog articles about your ideas you know so more people will see it more people will use it and since you have the uh support from uh the MCP uh AI tool uh it will be very easy to show it to show it also. And uh the finalists all of you you get three months of this alien gateway credits and of course you get also visibility. So you know being a finalist doesn't mean you didn't do something correct. It means you did something very good and the difference was maybe in the details but still we will support and we will give you the floor you know to to to improve to do something you know maybe that you didn't have the time to fi finalize or to check. So every everyone is a winner not only you know one prize winner everyone you you you have opportunities. So what we will do next uh by the end of October hopefully I understand that some of you may not have the time uh to do it uh you know next week or in two weeks. So we will uh contact by email and we will find uh common time slots and dates and we will have a call and then we can see uh how we promote your work uh when uh and then we can start seeing if you want this prototype uh to be able to be uh transformed to something more concrete. So now it was an idea a prototype. In some of your applications we saw uh some comments like uh I run this with a small uh data proportion because you know this is a prototype. So we can have now more advanced let's say trial, error, improvement, feedback phases and we have time to do it and also we can have a we can check also uh the maturity of the things you created. So can we at the end we need to ask you know after 3 6 months we need to see is the application mature enough. Can we integrate it? Can we uh can we maintain it in a way that the community will reuse it? Is it uh something that uh we should let's say place it inside uh one of our services or shall we keep keep it independent and then users can reuse it. So we need to think of these ideas based on what you created and uh have a plan have a common plan. Um that's it from me. I went very fast I think but you know I couldn't keep it a long time you know like waiting for all of you. We want again to thank you so much uh for everything you uh created. We will be here for you. We will support as you see uh as you s so 3 6 uh months 12 months. Of course there is not a strict deadline. You know if you don't do something in 3 months doesn't mean we will not be here for you. Um I think if if the winners are here maybe we can ask uh for them from them to say a few words. I need to stop sharing because I will have the chance now to see you all. And again community vote is number four. Uh as here theme winners 4 5 8 12 and the grand prize is uh data footprint number seven. So let me stop sharing please so I can have a look now uh to see uh all of you. Hi, >> I am the [clears throat] the winner of the the theme D. Do you hear me? >> Yes. Yes. Hi, Alexis. You are number 12, right? >> Congrats. >> Thanks. And thank you. Thank you all and congrats to to all. It was a really interesting uh subject. The kon was really great. Uh I only have a few minutes but uh uh left but uh thanks from from France. [laughter] >> Thank you so much. Welcome. Welcome. And um thank you because you know when I saw the theme it was financial right financial finances I was thinking wa this is the first time I see something like that. So uh I was curious also because you know most of the times uh the topics are more close to research topics and that was something new. So thank you for the opportunity and we will keep in touch now in the next uh days. Um is there any here maybe number four or uh five? Uh number five is Xan. So if Jean is here maybe we can ask I think Takahiro was here. Takahiro is the applicant. Tahiro number eight. Hi. Yes, thank you very much. Uh I'm very honored to receive the same three uh SIMC award. Uh I'd like to thank open air and area intelligence for this opportunity. Uh my research basically focusing on uh the quantum technologies and uh especially the emering technology including AI semiconductors. uh I focus on this uh hackathon uh especially the quantum technologies. So the open quantum evidence address asks the one simple question uh uh can we replace EU Japan quantum uh research from funding to publication and software. So the uh atas makes this uh auditable and uh uh reproducible and um same method can be extended to AI semiconductors or other emerging technologies. So I hope to develop the further with the open and uh community. Thank you very much. >> You're welcome. you know, it was a good idea because I remember when we I saw the demo the app that you had online. Uh I remember I saw the topic that uh you you scanned the information on the graph focusing on Japan and then you try to find out uh funding uh let's say paths and I I was thinking oh that could be great because we have a big community also uh on Japan and maybe we can really start showing this to policy makers could be interested for sure and people from universities because that gives them a a very uh you know transparent way and trustable way to understand how uh in the country how uh the funding goes where and what happens next. So thank you so much and >> congrats you know great idea great >> and maybe I I think Anna Gas I think number seven you are here right? I think I saw you. >> Hi. >> Hi. So congrats. Congrats. >> Thank you. >> I'm I'm very thrilled. Uh I mean I didn't I wasn't expecting that but I'm so happy. Um this is part of my uh researcher. I am I'm doing a PhD um focusing on the role that might play uh research data in research assessment and it's very important to have more detail and to have you know a way to identify um what research information is telling us about the data that is published and when I saw the hackathon uh it was an opportunity to explore and play and um well I'm really happy and I'm looking forward to see what else we can do with this. So, thank you again. >> Yes, thank you very much. And you know, I understand that you you had a personal experience. So, you wanted to, you know, apart from the research you run. So it was personal experience on research assessment plus your PhD your work and you combine them uh and now we have something very very good to work on you know very very good and I can share if you want I'm at the hackathon page and maybe I can I can copy paste now the in the chat you know we will send emails of course after the this uh call uh we can share all the documentation, the leaks, the code, the demo videos. So, people that might have missed it, they have an opportunity to go, you know, and check them out. >> So, that's really good. That's really good. So we have um we have people one of the things I liked it was because we have applications from uh Europe you know like you you and people like uh um Takahiro from Japan and we had people applying from Russia. So it was not just uh one country, five countries as I said it was many countries not only Europe out of Europe which shows also the potential of open science know how powerful it is and people want to see uh very very good results out of it and and explain why they use it how they can get the value. So I'm really I'm really happy we run this kind really also and AI because you know I have the the impression that it gets out of the ordinary hackathons we run so far because it was also AI assisted so I understand that you had extra uh help from AI is it true if anyone can say you know how you experienced also this new tool that you had in hands apart from graph of course which was there for many many years. Uh >> well I I I think the the AI help you to you know um to get easier for the non technicians to work on a workflows with the information. So um and for me a value of the of the project is that um anyone related to research information can use it and uh find a way uh to to you know the the researcher uh could have a better narrative curriculum. that um a data creator uh can see oh what is happening why citation is not uh is not coming uh what what which workflow we needs to to improve to see the information put okay all there so I think AI allows us the non techchnical uh or not so much technical people to to work with data in another way so it was it was a good experience Yeah, thank you so much. Uh yeah, I I feel it as a simple user also because when I try to find a paper or publication on the topic and now I use this uh alien AI tool uh that uses the graph and I found many many more information there and I was sure that whatever I get it's going to be correct. It's not hallucination, you know, it's not something that AI imagined. It's there. It's on the graph. I can test it out. I can verify. And uh one another thing, uh I was thinking of you know maybe ideas is that if we try maybe to you know make this a more concrete tool, not just a demo, a gateway that we have for hackathon. Maybe we can also offer a new tool to our users. >> Uh that they don't have to be technically with a technical background like new they can just go and search there. So it's really important >> and more and more ideas. We'll see. We'll see. Uh thank you so much Allah. Thank you so much. >> Thank you. >> And I have a feeling that Masimo that is here. [laughter] Hi M. So because you also applied. >> Yes. Yes. >> And it was a was very good opportunity for me um because uh as an independent uh developer as someone who is not into an academic background having the opportunity to participate to um as I work many years on new model architecture. Um I would really like to connect with more researchers also to validate uh the architecture of the pipeline as we use the uh AI tools like LLMs to uh basically um build an a pipeline for classification for a to classify papers >> and uh it was very funny um and um yeah was uh was very great to have this opportunity also to to experiment actually one of the first deployment of this kind of small models that I'm trying to build. So it was really great. >> Thank you. Thank you so much. I I remember you know uh the application it was uh it was it the case of rejection human if I reject something it was this case remind me. Yes. Okay. >> Yeah. Yeah. Yeah. Yes, there was also a possibility also for contributors to to add human evaluation. So that was not only AI. So trying to to put human in the loops. >> Yes. Yes. Yeah. >> You know uh we will uh we will of course contact all of you again. You know you you you created something. we will uh we can have a look more intensively and say these are the reasons you know uh we think you you could improve and maybe that's why you're not in the winners list but it's still fine you know we are not judging people here or ideas we want to test out what we can do what we can offer and to the community so thank you so much because every idea has its value every idea and every implementation you know Um thank you so much. Uh do we have any question from uh the graph community uh audience that normally follows this call and now we hijacked it with a hackathon. So if there's someone who needs to ask uh the teams uh anything please unmute. You're welcome to unmute. If you need to know something more, um, feel free, you know, because as I said, we will run now, uh, blogs, articles to present, uh, into more detail uh, all the applications. So you can have a a broader uh idea of what was uh delivered, how uh when is the code, where is the use case. In some cases, we might have uh the opportunity now to use a use case. For example, uh number four, if I remember correctly, the community vault has a use case, which means we can take this use case and we can apply, we can save it maybe as a training uh material on open plato uh the open air training uh platform. So uh people that don't know uh how to use uh all these AI tools they can follow this usea use use case that had many prompts and you can replicate and see how you can use prompts to find out uh information that it's valuable for you. It was um it was using uh a judgment if I remember correctly number four was the mathematical uh judgment. So it was checking also on from mathematical perspective uh which means good value uh trustable uh information uh the results uh of the prompt. So what you get back from AI. Um and just to mention we had also some ideas very good ideas also that we will definitely uh as I said uh contact owners of the applications of the code of the demo whatever it was uh that did one thing they didn't use the MCP AI at all in these cases we will go back and discuss has uh will have a new approach and uh you might see more in the future, more applications in the future uh from uh from this hackathon and more collaboration, more pilots. So you might see many pilots coming out of this call and you can use them because they are now uh well tested. Um anyone again a last opportunity please if you have any question for the hackathon or the teams uh the winning team is here so if you need to ask something okay so if if there is nothing new uh then we will keep in touch uh with the hackathon again through social media and then through uh future open air graph community goals and uh thank you very much all of you for joining and my colleague Alen that runs the call and my colleague Miriam from Italy from Senar that is the part of the opener graph team. So thank you. Thank you >> and thank you and thank you Andiki and to everybody who participated in the hackathon as well as the finalists and winners. Uh it was really really interesting to read all your submissions. So congratulations again and thanks again also for participating and uh we'll see you in the next community call or uh around via the help desk or use our forum. Uh we'll always be in touch. All right. Thanks everyone. Have a great day. SP.