Live AMA: SahAi for Shiksha Hackathon 2026 | Ask Mentors Anything 🚀

Added: 2026-05-22

[00:01:06]Hi everyone, welcome to official AMA ask me anything for the Sahai for Shika hackathon 2026 a mission by Vadwani supported by google.org and powered by reskllasi from reskll and I'll be your host for today's session. So before we start, if your teammates couldn't make it, don't worry. We are recording this session and it will be available to you shortly after the recording. This hackathon is a huge opportunity to learn and build AI that actually works for India. Our experts are here to help you make your ideas better. So please ask anything. No doubt is too big or too small. So joining me are our mentors from Vadwhani AI. Firstly we have Vishal Naidu, senior program officer education. Then we have Sriat Mukharji, lead pedagogologist.

[00:01:56]Then we have George Tom, associate ML scientist and Anika Murka, associate ML engineering. And joining us shortly for Q&A will be Weber Yadav, senior product manager. So before we dive in, a quick reminder, please keep your questions coming in the chat. We have a dedicated Q&A segment at the end. So now let's get started. I'm going to hand it over to Ashal to take you through the mission and the hackathon journey. Over to you.

[00:02:27]>> Thanks. Uh do I have the tech shared?

[00:02:31]Yeah. Hi. Hi everyone. Thanks for joining. I'm Vishal uh senior program officer at Madwani AI and excited to have you all uh on the session. Uh so this is a 60-minute session and this is what we want to cover today uh in terms of who we are and why we are doing this hackathon and what is the overall journey of the hackathon and what are the problem statements and what should be the considerations while you are solving for these problems and then uh we'll deep dive into what are the expectations out of the phase one of the hackathon what would be the evaluation um criteria and also then what would be the phase two uh journey would be looking like uh and then we'll have we'll end it with time for Q&A where you can uh have your questions. So as we go through the presentation feel free to drop your questions and we'll take them through the Q&A session.

[00:03:24]So yeah uh a bit about Wadwani AI. Uh we are a notfor-profit applied AI institute uh which builds AI uh solutions and for social impact across health, education and agriculture. uh we essentially design AI solutions for public uh Indian public systems. So so far we have deployed around 25 plus AI solutions across these three sectors and we have a core uh a IML team inhouse uh led by our uh chief scientist Dr. Alman and these are some of our donors and this uh hackathon is uh being uh run under the uh grant and initiative of the Google.org mark that we have for the education. Uh so why this hackathon right uh so there there are a lot of real uh world and uh complex challenges that uh the indician education system has and many of these uh problem statements with the advent of AI and the way AI has been progressing it could be uh solved uh and it AI could actually be leveraged to solve some of this complex solutions and uh and especially when it is uh designed thoughtfully and uh applied in ground reality. So this hackathon is uh aimed at bringing that uh or catalyzing that innovation to build AI first solutions for some of the real world complex education challenges that we face in Indian education system across early childhood school education and higher education. So it is informed by ground reality and designed for impact and scale. And what this hackathon is definitely not is not just a demo competition but it is essentially a platform to build solutions which could be pilot ready ideas that can really land in an Indian classroom.

[00:05:18]Uh so what would be the journey looking like right and the hackathon has been launched on 15th May. uh the registrations are already open and today we are around 22nd May doing the session and the deadline for phase one where you pick up a problem statement and submit your ideas is on 3rd June uh and 2 days after that we'll be shortlisting the top 30 teams and we'll be announcing the top 30 teams by end of uh 5th June and right from 6th June to 20th June which is a 14-day online build sprint which is your phase 2 where uh the top 30 teams will have mentors mapped to them and the mentoring framework everything we'll cover in the upcoming slides but uh all these top 30 teams will then develop a functional uh uh MVP by end of this phase 2 uh and which is 20th June and within these top 30 teams uh the top 10 will be shortlisted uh and we'll announce them by 23rd June and these top 10 teams then will fly to uh New Delhi uh for the grand finale on 3rd of July where the top three uh would be rewarded uh uh based on the jury's evaluation and again for all the details would be covered in the upcoming slides. So eligibility who can apply uh it's open for any individuals either students working professionals or researchers and uh the team minimum team size is two members and you can go up to five members per team. uh an ideal number would be four member per team and the registrations and phasement submissions everything would close by 3rd June uh 11:59 p.m. IST and uh the we have three tracks across early childhood, school education and higher education and together within these three tracks you have 16 problem statements uh and we'll go deep dive into those problem statements in the upcoming slides and as I mentioned the grand finale is on 3rd July and if you are a solo applicant uh feel free to register and join the WhatsApp group you'll have a team finder sheet on the WhatsApp group where you can list your skills and uh there similar uh people uh with similar or complimentary skill sets on the group who are also looking for teammates and uh you can find your collaborators.

[00:07:37]I'll hand it over to Srijit our lead pedagogologist to discuss about the problem statements.

[00:07:48]U we are uh looking at three tracks and 16 problems across the three tracks and um I'll take you uh one by one but the three tracks of course are early childhood uh education, school education, higher education and skilling. Uh in the following slides we'll just deep dive into Yeah, thanks so much for changing the slide. We'll just deep dive into each of these problem statements and then later on in the question answer sessions we can uh take any questions on each of these uh solutions or slides. Yeah. So the first track is early childhood education and the first one is access for vulnerable children. Now we know that uh we have a huge population a growing population which goes into primary schools. Um a lot of children are not fortunate to grow go to primary schools and there are a variety of reasons. One of them could be that they are children of migrant workers who keep on shifting location.

[00:08:51]Some of them uh are tribal and not soioeconomically well uh well off to afford schools. Um >> so what happens is that there is this gap in access to education. there is this gap in continuity in education and uh we we are not able to sort of get every child uh a decent uh education.

[00:09:14]Right? So the first problem is how do we improve access for vulnerable children um they are like I said children from the migrant tribal communities and socioeconomically diverse or hard-to-reach communities. That's the first one. Uh the second one uh is the quality of play-based uh early learning.

[00:09:37]Through theory, through practice, through evidence, we know that play-based learning is a good method for children uh especially in the early years to learn, right? But we don't have good quality. How what can we do to have good quality and deliver high quality um play-based learning experiences?

[00:09:57]We we know that frontline workers have a lot of duties. they might not have the support to facilitate engaging early learning experiences every day. So that's the second problem. So how how can we solve this problem of not having quality play-based early learning?

[00:10:15]Moving on to the third one is the holistic child development tracking which means that uh based on the new education policy a child needs to not just learn but learn holistically which would include uh you know a socioeotional quotient uh an ethical quotient a cognitive literacy aesthetic cultural physical quotient and we need to sort of synthesize get a process in so that all of these metrics throughout about the child's learning cycle is tested and tracked. So the third solution is to solve for tracking all of these different different uh quotients or all of these different different um holistic development metrics for the child. Uh and then the fourth one is uh to the Anganiari worker. The angiari worker uh is very very stressed is stressed because they have multiple duties. Um they don't just look after education, they look after nutrition, they have home visits, they have to report a bunch of things, maintain registers for a bunch of things uh and report to their supervisors. So in doing all of these things they are and meaningful child engagement. So what are the things that Anganari workers can uh be supported on? That's the question.

[00:11:48]That's the that's uh what we are going to tackle in this fourth problem. And then we have the fifth one like early identification and referral of children.

[00:12:00]If somebody shows dis uh uh slow learning signs, if somebody shows developmental delays or disabilities, how can we report them quickly so that they can be segregated and given their level of education? They can be remediated, they can be managed. Um so that's that's a problem which we often ignore. It is also something that parents don't want to acknowledge, teachers don't want to acknowledge steadily. But that is a real problem and if you don't do it right at the early uh childhood education stage, these problems can grow. Children can fall back in classrooms and uh they can they maybe eventually drop out or get disinterested. So that's that's the fifth problem.

[00:12:50]Uh then there is track two where we are talking about school education and the first component is inclusion and classroom participation.

[00:13:00]Again this is similar to the uh problem in the previous track where there are within schools there are children who are coming from migrant families, tribal families, disadvantaged groups economically and socially weaker sections. So how can we ensure that they are participating fully in the classroom they are learning as well as uh their peers who are not from disadvantaged groups and um you know school teachers don't often have uh remedies don't often have hacks to treat them to teach them equally um so what can we do for these children so that's the first one under the school education track the second one is foundational literacy and numeracy. We have had a lot of emphasis on foundational literacy and numeracy in the new education policy 2020. And the real deal is that if you don't have good foundational skills, then you'll not be able to um do well as you uh grow up and go up the uh learning ladder to uh you know to higher grades. So if your foundations are right, if your basics are clear, only then you're going to be able to succeed in your future grades.

[00:14:14]And then people often say that you know you're in grade you're grade 10 arrested when they were in their for of also. So that is the next uh section in this uh track.

[00:14:37]Third one is critical thinking and future ready skills. Um with AI critical thinking has become very very prominent as a domain. Um we want to see that you know uh people uh children are not learning by wrote they are learning about um concepts and using those conceptual understandings, applications, problem solving skills etc. um so that they don't have you know limited opportunities for their careers they can think outside the box and they can think the in the right direction so that is something that we want to cater to also uh within the 2.3 of the school education then the fourth one I think is a very critical one is uh learning gaps and timely feedback right teachers often don't have the data teachers often are not able to identify y the learning gaps um and since they are not able to do that they are not able to provide targeted interventions to close the learning gaps. So this is a critical piece again where everybody uh within the school ecosystem should be working towards bridging the learning gap so that the child who's you know who's coming at the top of the class and the child who's lagging behind the gap in that learning can be closed. Uh then we have the next one uh which is teacher professional development. Uh we talk about teacher training. It is a very very essential component but uh sometimes it happens only once a year and that does not get translated. Uh teachers feel disconnected because there's no follow-ups. They don't feel that um the the the trainings are enough. they feel that they need professional development that is continuous which is practically translatable to their classrooms and um you know which are a part of their context their um their environment. So that is something that we also want to tackle in under school education and uh finally uh the school uh decision making and early intervention is is something that has come up in a big way.

[00:16:53]So school uh and education systems they lack integrated decision support systems that connect student learning uh attendance assessments and administrative data in which uh you know there can be uh an immediate and um timely planning and intervention. So resultantly what happens is that teachers and school administrators and school leaders they struggle to identify risks at early stages. So if we can have an integrated platform which tracks their assessment the students assessment data their attendance etc and give a comprehensive report that is something that will be very very useful uh for all the stakeholders in the schools.

[00:17:40]Uh next we move on to track three. In track three, the first uh one is access to skilling and career pathways. Again, learners from rural, low-inccome, first generation and socioeconomically disadvantaged backgrounds, they face barriers in accessing quality skilling opportunities. Um so what can we do about them? Um these barriers basically they they show up across uh the funnel in terms of low environment in skilling programs, higher dropout rates etc. So we need to arrest this problem also. So what can be done to arrest this problem?

[00:18:16]Uh moving to the second uh one in this track is of course job readiness and entrepreneurial skills. Many learners complete programs without developing the applied and sort of industry relevant uh entrepreneurial capabilities.

[00:18:32]So uh you know this has two sort of tracks to it. One is that they need to be employable. If not they need to be skilled enough to begin self-employment journey. So how can they do that? So that's the next track in this. The third one is uh skilling assessment and professional signaling which is that uh you know assessment systems need to be robust enough to capture what learners are actually able to do and not what they have you know read in their coursework. So because there are few fewer and fewer of these kinds of assessments this it it it makes it difficult for employers to identify skill gaps and guide improvement for potential employees. So this is something that we also need to work on from a from a uh skill assessment uh point of view. And the next one here is uh faculty and trainer capacity. So this again is capacity building related where often uh trainers they lack timely support and the tools. There is no feedback loops to update content um and teach applied skills effectively. What can we do or how can we build a system which uh um support you know uh teacher train for faculty training uh and teacher capacity in the skilling environment. Uh fifth is to have education to employment pathways. Right? So learners often face fragmented journeys across institutions in the sense that um they don't have proper career guidance. They don't know what their assessments is going to really help them do. they are also you know they don't have enough opportunities or information um about apprenticeships about campus placements for example and uh lifelong learning opportunities. So this makes transitions from an academic sort of setup into a working sort of setup very very confusing and inefficient and we need to support our learners in that. So this was the fifth uh fifth track within the higher education and skilling track. So like I said in the beginning if you have any questions we can take them later.

[00:20:56]Over to you Vishal.

[00:20:58]>> Thanks. Thanks Rajat. Uh so yeah that's the overall uh tracks that we have under the problem statements and just wanted to highlight uh one thing that while uh we would want you to choose a specific track uh that's a mandatory thing you can't uh build solutions that span across the three tracks you have to pick any one track and even within the track the ideal scenario would be if you're picking one problem statement but if you think your solution is scattering to multiple problem statements within that track that should be fine. But yeah, but sticking to one track is something that we would require you to do. Under the track, you can pick uh ideally one pro one problem statement or if you think your solution is scattering to multiple uh problem statements, it's okay to do that as well.

[00:21:45]And uh just one last thing uh while you start building on these solutions as your uh uh from your problem statements is as mentioned in the initial slide we are we want to build solutions that can create real impact right and this is we are targeting this for underserved communities. So if you while you start thinking about the solutions we would want you to keep these three pillars in mind which we are calling the design principles. uh but essentially first pillar would be the technical robustness and real world readiness. We prefer uh to have offline ready and low connectivity solutions. But again we understand that every uh solution can't be offline ready but think in this direction on what we can do so that it's also serving the last mile at in any part of the country and especially the underserved communities that we are targeting to address this work. And you can also think of hybrid solutions where you know it could be a sync mode where it's not completely offline but partially offline. So think of be innovative in terms of building the solutions. But this is a preference that we would have while you're thinking of the solutions. Uh and the second is yeah think of the compatibility with uh entry-level smartphones and low compute environments. And also be thoughtful and build solutions that can handle uh fragmented and inconsistent data from live systems wherever it's applicable depending on the solutions that you're building. And we would prefer you to uh build based on open source or costefficient model choices uh as you pick up your AI models or design your AI models around the solutions. And the second pillar being the human- centered design. So how do you uh think of supporting regional languages and local dialects wherever relevant. So the inclusion part is very uh much required and if you had even identified in the problem statements each under each lag the first problem statement was also targeted at uh solving for inclusion. Uh so that is a very big piece for us and also most importantly think of solutions and solve for sol build solutions that will augment teachers and system actors and not just uh you know not aimed at replacing them but actually aimed at supporting them and augmenting them and yeah as I mentioned earlier design for accessibility diverse learning needs and diverse learners and the third part is feasibility scalability and pilot readiness. So here put yourself in a hypothetical scenario while you are uh final building the solutions that it is not just for the demo but uh think uh if this solution is would it be uh can we pilot this in a district or a real district or an institution within 12 months that's the kind of quality that we are expecting at the end of this uh funnel of the hackathon. So while you're designing for solutions think on in such a hypothetical uh perspective and also whenever you're building uh think of modular architecture which could be replicated across states and different contexts and also most importantly it's not just the technical build but also think of uh who would be the users how would they you know how will you get them to use the solution and what how would you integrate into real systems so the adoption plan and stakeholder mapping is something we would want you to think through as you start uh building the solutions. So now I'll hand it over to uh George to discuss about what are the expect specific expectations of the solutions uh from phase one.

[00:25:24]>> Yeah. Uh so the >> Yeah. Yeah. So for high so for the phase one we have uh four deliverables and the red for the phase one is on uh 3rd June midnight. So the first one being the solution title. Uh so uh give your solution a title and you know give it a catchy name and the second one being a clickable prototype that ideally uh is in Figma or in ADOP or you can even uh use Google stitch for this and for the third one is a video walk through uh which basically walks through the user flow and shows basically explains how it works. You could either upload uh upload the video through YouTube as an analysted video and share the link or upload it to Google Drive and then share the link. And for the last we uh we look for a presentation tech which for in which we uh require like 10 sections.

[00:26:13]Yeah. The first one being again the problem statement uh uh why was it chosen and you know what's the evidence behind it and the second one being uh who are the target users and how are they going to be you know uh how is the solution going to help them and the third one being a solution overflow and again like a a key workflow of how it's going to work. Fourth point being why is AI necessary in this particular solution? Uh what happens when you don't have an AI? Why do you necessarily need an AI and would it fail without it? So basically answer these questions in point in section number four.

[00:26:47]And next is an AI methodology. Basically what is the choice of the model and what is the reason behind that? Uh and for the data like do you do we have data available? Do we need to annotate data?

[00:26:59]Is it is it feasible to annotate such a data? uh what is the evaluation plan for those particular model right and uh moving to next part you know we look for like a high level system design architecture on how we going to build it and the next one being again uh since most of the solutions are going to be in you know deployed in government schools like what is the real world readiness we again like uh we shall explain like you know we look for like models that preferably work in offline setting low connectivity and also support based on the solution also support like you know multilingual languages and regional dialects and eighth one being the imple implementation plan on how we're going to implement it the milestones dependency what challenges do you think you're going to face in this and uh ninth section being the team composition and the skills and these are the required sections apart from this I I mean feel free to add more sections based on uh the you know the problem statement or how that you're going to pick yeah uh now looking at the rubric uh on how we going to uh score the phase one submissions we have seven criteria the first one being the AI approach and technical feasibility where we have a 25% weightage for that so yeah the next one being uh uh equity inclusion uh responsible AI and again the multilingual consideration so uh is is your application going to be fair for every users that is going to use it right and that has a weightage of about 20%age uh and for the next uh with 15% weightage we have the pro problem framing the user understanding and the evidence on you know if the solution is actually worth pursuing or not.

[00:28:35]Next one being the solution design and prototype quality which which carries over a 15% weightage and uh is your solution uh an innovative solution is it is it new are there you know any existing ones or are you augmenting some existing solutions based on that there's a 10%age weightage and the next one being the impact and the scale potential uh which carries 10% weightage and uh last with 5% weightage is implementation plan on Yeah.

[00:29:08]Yeah. Uh so for the phase two uh how phase two is going to work an is going to walk you guys through uh the phase two section of the uh hackathon mentors. There will be multiple panels uh where each panel will contain four people uh a domain expert, a IML expert, engineering expert and a product design expert. Each a shortlisted team like after phase one and each team will be assigned one panel. There will be two ways you can contact your uh mentors. Uh there will be a mandatory sync that will happen in the middle of this 14-day sprint. just a basic check-in where your mentors can understand where where you are at the solution and you can ask some some uh questions that you have in mind with them. It will be a 30inut session with all your mentors and then there will be a way that you can asynchronously also reach out to them.

[00:30:16]Uh you will be having your WhatsApp channels with your uh mentors added to them. You can ask questions from them.

[00:30:22]You can also ask if they are free they you can have a session uh planned with them like a 15 20 minutes quick call with them if you're stuck anywhere. Um so yeah so the mentor will be there to guide you through the process for this phase two. U moving on yeah so like if uh basically this will this is how the time phase two timeline will look like.

[00:30:45]It will be a 14-day build sprint where somewhere around on 11th you'll have a mental like u till 11th we'll try that all the teams have a check-in with their mentors so that uh they are they're not stuck in the initial few days of their uh MVP delivery and obviously like the all 14th days you can anyway like reach out to them asynchronously and uh yeah so on the 14th day these are the things that you have to submit like the functional MVP which is end to end user flow integrated with your AI component a documentation a working demo what if you have done any evaluation those evaluation results and what will be your plan for deployment and sustainability of your solution.

[00:31:33]Yeah. So yeah this is the most interesting part I guess. Uh these are this is the price pool. We have a 15 lakh prize pool for the winners and like six lakh for the grand winner, four lakh for the first runner up, three lakh for uh the second runner up and then we have a consolation prize of two lakh.

[00:31:51]Uh the top 10 teams after phase two will be u will we'll have offline the last finale we'll have offline on 2nd July I guess. Yeah, 2nd July in Delhi. Uh looking forward to meet all of you there the finals finalies. Thank you.

[00:32:10]>> Great. Thanks. Thanks, Ana. So, yeah. Uh just uh the finale is on 3rd July. Uh and uh thank you so much George and Anika for taking through the details of phase one and phase 2. Yeah, just to summarize uh from phase one, we'll shortlist for phase one uh uh after 3rd June uh we'll shortlist the top 30 teams. So you have 12 days uh for your submissions right now and from these top 30 teams uh under mentorship they will be building the functional uh MVPs and working demos and phase two and within these top 30 teams uh after they build their MVPs we'll shortlist the top 10 who will go to the grand finale and from there we'll uh the jury will pick the top three teams. Now uh currently yeah as I said we are 12 days away from uh the phase uh one deadline. So form your team. So we would suggest you prefer you know having a ski team which has mixed skill sets A IML engineering design and also if you have seen the stress or the weightage that we have given in terms of understanding the real world scenarios and the education uh systems. It would be ideal if you have someone in your team who knows education while you'll also have mentors from the education domain for phase two even for your phase one build. that would be a good uh way to have uh a balanced team. So again uh feel free to join the WhatsApp groups and uh you know you have a team finder sheet there fill in your details and even if you are from an education background would uh definitely suggest you to participate in the hackathon because there are teams that would need your inputs as well and the other way around and yeah from the sheet you can still uh find your match and you can form your teams and then yeah pick your tracks and the problem statements and start uh look at the design principles.

[00:34:08]once again before you build it and yes uh that's and yeah just to summarize I think uh repeating the same point the strongest teams uh pair tech talent with someone who knows education from the inside be it a teacher or an education professional or somebody from the edtech or education domain so that's what separates an idea that works uh from an idea that just demos yeah that's what uh we wanted to share from our side today uh So you have a website where you can register vadwani aai.reskill.com.

[00:34:42]Uh and on the website you'll also have the WhatsApp group. I'll also I think it's already there on the chat as well.

[00:34:48]Request OJI to drop the WhatsApp group link once again. You can find your team finder sheet there. Fill in those sheets and uh form your teams. And yes, you'll also have this recording. Uh now we can open up for Q&A.

[00:35:05]>> Yeah, sure. Thank you team for the deep dive and I see that we have had some great questions coming in. So let's bring them up one by one. Swaditi asks what are the judges looking for the most the idea or the execution?

[00:35:26]>> George uh would you want to take that for phase one?

[00:35:30]>> Right. Right.

[00:35:33]Yeah, we do look at both actually like so we want a solution that we would be able to you know at least pilot I uh at least by phase two you we would want the solution that we would be able to pilot uh maybe you know by end of the year or something. So we want solutions that are realistic and are possible to build upon. So we give uh I don't in a way we could say that we keep similar weightage we want we going to we the idea and also the execution.

[00:36:01]Yeah.

[00:36:02]>> Okay. Thank you George. Vishal, you want to add on to that?

[00:36:07]>> Yeah. Yeah. And as George mentioned, definitely the implementation is also really important. Essentially the implementation feasibility, right?

[00:36:15]Not the execution of how we are building it. that you know how feasible if this this idea that could be implemented in the scenario or the context that we are looking at especially the underserved communities and whatever we have discussed in terms of the design principles that's uh also that also has a very uh equal weightage as George mentioned yeah >> okay so the other question is is it necessary to have at least two members in a team Yeah, we would uh we would want to have at least two members in a team as we said like it's a collaboration and we'd want to have mix of skill sets while you are building for it and again uh reiterating that you will have there are a lot of uh people in the WhatsApp group who are looking for teammates. So if you are a solo applicant you I'm pretty sure you'll find a teammate once you join the WhatsApp group and also feel free to share it with your friends uh who are in different domains and see if they I mean we would love to have more and more people participate so that we get the best of the talent to bring up the innovation.

[00:37:26]Yes. And also guys u if you check WhatsApp community uh we have linked the pinned the uh team finder sheet. So you can find uh team members according to your requirements over there.

[00:37:42]So the next question is does the prototype need to be working offline?

[00:37:47]>> Uh I'll take this first and uh maybe then Anika or George can also answer.

[00:37:52]The idea is uh it need not work offline at least the phase one we are just looking at a clickable prototype it need not be a working prototype it's more of a design flow of uh how your solution would look like uh in phase two is where we are looking at a functional prototype uh that would work but in uh definitely for phase two as well as I said we would prefer to have solutions that might work offline or you know that can work in an offline setting as well it need not be completely offline it could be solution where it could either be syncing data uh when they have internet access. But again, these are all preferred uh these are preferences. But if you think you have a great idea and something that can't be built offline, feel free to do that as well. But yeah, consider these uh uh options while building the solution. That's our only uh take on this. Ana, do you want to add anything to this?

[00:38:48]No, I think yeah that much that pretty much covers all the points that I have in mind.

[00:38:52]>> Yeah, great.

[00:38:56]>> We can move on to the next question.

[00:38:59]>> So, can grade six to seven students also participate in this?

[00:39:06]>> Yeah, sure. We we are open to anyone uh who would like to participate in this.

[00:39:14]>> Okay. So Panch asks if we need all of these five statements in a solution or only one.

[00:39:21]>> Uh as we mentioned uh it's basically the track is mandatory and we would ideally prefer if you pick one problem statement under a track so that there is more depth to the solution. But if you think your solution is solving for multiple problem statements within that track uh it's uh okay to pick up multiple problem statements. But yeah ideally I mean in a normal scenario we would want you to pick any one problem statement with the network.

[00:39:49]>> Okay moving on to the next. So Anul asks can our solution address both school education and early childhood or do we have to pick one track only?

[00:40:00]>> Yeah I think answered in the previous one you solution has to be fitting any one track. Even let's say your solution is solving for both would want you to pick any one uh track decide which track it is most efficiently serving which uh target group it is most efficiently serving uh or solving the problem for and would stick to that track is what my session would be.

[00:40:26]>> Okay.

[00:40:28]So the next question is if do we need to show a working demo in phase one or is the video walk through okay?

[00:40:36]>> Uh George you want to take this?

[00:40:38]>> Yeah. Uh so we don't expect a working demo for phase one. We do expect that for like phase two. Uh but for phase one we what we expect is a clickable demo uh which is on Figma or uh Adobe XT and a video walkthrough of that uh prototype.

[00:40:58]Yeah. Okay. And design just to add to George, it could be any other design uh software as well. So whatever you think you're comfortable, we just want to know the flow of the solution in ter from a design aspect. So we are not really uh you know uh there is no requirement on the tool that you would use to create the clickable prototypes.

[00:41:19]>> Right.

[00:41:20]>> Okay. Moving on. Shoubam asks where can we access scoring rubric?

[00:41:27]>> Uh it's on the website. Uh while you'll also have this recording, it's also on the website. If you click on guide on the top right uh home tab on the website, you'd be able to see an assessment section where you'd have the phase one scoring rubric.

[00:41:43]>> Okay.

[00:41:45]Um Kiran is asking the solution that you are expecting is an app or like a web dashboard.

[00:41:54]>> Ana or Josh, you want to take this?

[00:41:57]>> Yeah. I mean it could be Yeah, it could be anything. I >> could be anything, right?

[00:42:03]>> Yeah. Yeah. Like for phase one, it is just a prototype. For phase two, it could be any working uh demo of app or web dashboard.

[00:42:16]Okay.

[00:42:18]So now uh can we submit the working demo in the phase one itself in the place of Figma?

[00:42:28]>> Yeah, I think you can. We uh I mean that what we have as a minimum requirement but uh if you are building more than that it's okay. uh but we'll still evaluate it from a design perspective and not from a working demo perspective to be fair to every other participant.

[00:42:46]Uh please add an or George to this.

[00:42:53]Uh yeah agreed like it you should submit the at least the basic things like a presentation a video walk through everything and instead of a Figma prototype if you have a demo it's fine but uh yeah I think probably like in the second phase then more features to be added to that demo right >> so do all my teammates need to be from my company or we can be mix of students working professionals etc. >> Uh it's up to you. There is no uh uh requirement or criteria about that. And the only thing I wanted to highlight this is you we would want uh participants to be registering individually and not representing any companies as part of this hackathon. So yeah, if you are a group of individuals uh who are colleagues as well, so you can join a team and participate. It's completely fine.

[00:43:51]Okay.

[00:43:53]So what does track agnostic innovation mean?

[00:43:58]>> Uh so basically the idea was like you have top three grand prize the first uh winner the first runner up and second runner up but in the remaining seven solutions that are being the seven teams who would be presenting in the grand finale if the jury finds. So this is a discretionary reward. If they find something that is really innovative and something they want to still reward for, that's the special price that we had. So other than the top three, if there is a special uh innovation that really needs to be uh rewarded. So that's when the it's on the discretion of the jury to either give this or not depending on the solutions.

[00:44:41]>> Okay.

[00:44:43]So for problem statements 1.2 and 1.4 there are existing applications like portion tracker used by millions of angani workers for nutrition and play-based learning videos.

[00:44:56]So okay I think that's the question.

[00:45:02]>> Yeah. Yeah. So I think if you look at it that way for every problem statement you might find some solutions out there but the problem statements are still existing. So uh that's how I would look at it and maybe others can also add uh so in terms of how you would want to innovate and see what are the existing gaps in the current uh status quo of the solutions or or of the systems and see how we can solve for it. That's the approach that we would look at.

[00:45:32]>> I think he has a follow-up question as well that should our solution talk this existing application for the solution or think from ground up. Quotion tracker has its own challenges and I'm not sure if you're building solution to replace it.

[00:45:47]>> Uh great question. Uh >> and I think more I'll just >> Yeah. Okay.

[00:45:56]>> The same problem goes for problem statement 2.6 as Vidya Samsha Kendra exists at national level and in at least 12 states. Same question. We build on top of this or an independent solution.

[00:46:09]uh yeah maybe I'll share my take on this and then I'll also request others from the team to share. So and uh this is again your uh team's uh decision in terms of how you want to uh solve for this in terms of either you want to build an independent solution if you think that is more feasible because we are also looking at the implementation feasibility and if you really think uh since there are existing systems and not building a parallel system is not suggested. Uh so if you think you have an idea or a uh solution that can embed or integrate into the existing systems yeah I mean that's also something uh we would be open to uh and uh if you think that is a more efficient way to build uh or solve for this problem then we want you to go in that direction as well and I'll pause here for others also if they have any comments on Sri Chut or others also if they on Um yeah like I also have a point uh like I think it will be interesting to have these findings in your presentation like the first phase is about ideation and ideation should have these groundings uh like why why do you why are you making this new solution if something is already there and it's interesting if you can actually specify if if you can if you have seen any uh gaps on on those applications and uh and how your new solution could be easily integrated into it.

[00:47:45]>> So like add these findings in your presentation if you already have them.

[00:47:53]>> Yeah. Ver uh do you also want to add this? Yeah, I think both of you have given valid points that the solution that any team ends up creating should be cognizant of what already exists and should innovate on top of that independently or using the infrastructure of the application that is already present. But we leave that up to you uh and you know bases your own evidence finding your own secondary research. If you're able to identify that there are potential areas where your solution can be plugged in, I I think that's a valid point. Um, but in terms of innovation, I think we're looking at something that can be placed on top of an existing application and not just replacing. Nowhere are we mentioning that it has to replace any existing application.

[00:48:43]>> Yeah, just I think we can move on.

[00:48:45]>> Okay.

[00:48:47]So moving on, Burun asks, I'm an entrepreneur and student too. So can I use the created solution for my startup too?

[00:48:59]>> Uh I'll again go first maybe and then I'll request Viber to also come in. Uh yeah, I think you can uh so because the IP will also be uh with you as you submit anything in this hackathon. So you should but it won't it again the only idea is it should not represent if this is already registered as a solution and all the idea is we don't want uh we wouldn't want any companies to participate in the hackathon. So that's the way we are looking at it. It could you could be an entrepreneur but if you are building a separate solution as an individual participant uh we would be uh that would be great but if this is an already existing solution and it is already registered as a solution elsewhere uh then uh we would ideally not want it to be part of the hackathon uh but this is my v you want to add to this or anything >> yeah that's correct v I think what we are trying to do here is striving for innovation and in terms of that we would want something that is built during the phase of this hackathon um and not something that is already existing in terms of a startup or a company. But if a submission in the hackathon turns into a startup, that's a great uh achievement if you're able to do that. But nothing existing should be reused here is the is the mode that we're trying to portray.

[00:50:21]>> Yeah, we can move on.

[00:50:27]>> Okay.

[00:50:30]Um can my solution be used on device compute on device compute?

[00:50:36]>> Yeah. Uh yeah your uh your solution can use on device comput and uh that's what we mainly mean by the offline uh setting. uh but do think about like the target users they uh I mean if you build a solution that really works with like flagship devices then and your target users do not have access to such high-end devices then that might be a problem and uh it's safe to assume in most of uh the cases uh the target users would ide you know most likely have a low-end device so do keep that in mind >> okay I'm from Jakan and I am building solution Can I pair up with my father?

[00:51:15]>> Absolutely. Uh there's nothing stopping you from doing it and we would love to see what you'll be building.

[00:51:24]>> Okay.

[00:51:26]Can a class 12 student participate?

[00:51:29]>> Yeah, we already answered this. Yes.

[00:51:34]>> Okay.

[00:51:38]>> Where can we get get data from? like Anganadi resistors were mentioned.

[00:51:47]>> Again, this would be very specific to the problem statements and you would have to do your secondary or primary research while you are building for it.

[00:51:54]There are uh a lot of open source data that's already available and also uh you might uh so which you can get through some secondary research uh and you might also have to do some primary research for some specific problem statements to understand the users and the context as well.

[00:52:12]Others feel free to add if you have any other points to this.

[00:52:20]>> Cool. Uh we can move on this. Okay.

[00:52:26]Is that fine the Figma screens having the version that what we are trying to do and also any standard temp template to prepare PPD or doc you will give any standard template. They're asking for a standard template. I think >> uh this is for the PPT is the standard template.

[00:52:47]>> No, I think essentially we want you to have the sections covered in the PPT. uh you are free to design it the way you'd want to but we would want you to cover all the sections which George has mentioned on his slides. Um and the first question I didn't understand it if any other anyone else could understand it feel free to answer.

[00:53:19]No, I think it's mostly to understand if there's a standard template that exists.

[00:53:23]But um I I think we've left the freedom to the participants to design any template that is necessary covering the sections as Val mentioned.

[00:53:32]>> Yeah.

[00:53:33]>> Can move on.

[00:53:37]>> Um what we need to explain in the video is it to explain our idea in video or how the Figma design will work in future prototypes.

[00:53:46]So we want you to once you build that Figma design or the clickable prototype we want you to go through this I mean it's a screen recording of that flow in terms of what are the solution uh what is the flow and explaining that flow as uh you record the screen and you can have a voice over in terms of what the screen does and what's what are the key user flows on that uh prototype.

[00:54:13]Is that clear? and Vber feel free to also come in for this.

[00:54:19]>> Yeah, that and also I think the video should be cognizant of how the user will be interacting with the prototype. Um what are the points of inputs from the user? How is the user transferring from point A to point B and what is the intent from transferring to point from point A to point B. So it's a advanced version of the prototype where you've been given an opportunity to explain how the user interacts with the prototype.

[00:54:44]So I mean that's what we are expecting.

[00:54:53]Okay. So we'll take last two or three questions now.

[00:55:02]So uh how should we create submissions for phase one? Do we need a full working product?

[00:55:08]>> I think this is answered. Again uh the requirement is just a clickable prototype uh which is more from a design angle.

[00:55:19]We can move on. Okay. Yeah. I think uh we will be circulating a form right in which you can just add those links after uploading to the drive right that's how we're submitting.

[00:55:33]Vishal you can correct me if I'm wrong.

[00:55:36]>> Sorry. Uh which question are we answering?

[00:55:39]>> The previous question they asked that how do we create submissions? So we will be circulating a form in which they can add the >> Yeah, it's yeah it's on the website. So it's on the website if you register you could see uh submissions where you can submit it and there you have the uh input tech I mean place the text where the form where you can actually submit your solutions including all this four spec parts that George has mentioned. Uh so all of that's there on the adwani.reskill.com.

[00:56:11]So you can look there.

[00:56:15]>> Okay. So it's okay to use edge compute using models like gamma 4 or light RD framework.

[00:56:24]>> Yeah, similar to one of the previously answered questions. Yes, you can use it edge compute but do think about the devices that the target users are going to use and see if uh those devices would ideally support these frameworks.

[00:56:39]>> Okay.

[00:56:41]Does uh our AI solution have to be a mobile app or can it be a backend backend tool or workflow?

[00:56:50]>> I think this has al also been answered.

[00:56:52]It could be anything any working solution.

[00:56:56]>> Yeah. But the only caveat here is that it needs to be presentable in terms of a prototype and a video. Uh whatever that uh includes I think it should >> Yeah. I mean yeah it should have some UI of course. Yeah, let's take >> Okay, so I think we have pretty much covered all the questions. So, thank you team for that deep dive and a huge thank you to all our experts for their time and for all of you for your brilliant questions. And if we didn't get to your questions, please don't worry. You can reach out to us through our support channels. We have uh shared our WhatsApp group link with you in the chat. So, please mark the deadline. It's on 3rd June. Only 12 days to go. So please start building.

[00:57:42]>> Anything to add Bishal?

[00:57:45]>> No thanks. Thanks everyone. Thanks for joining and thanks for moderating this and thank you Anika George and Viper and looking forward to the submissions everyone. All the best.

[00:57:57]>> All the best everyone. Thank you for joining.

[00:58:01]>> Thank you all the best.

[00:58:05]>> All the best.

[00:58:07]Thank you everyone.