Science AI & Innovation_ India–Japan Collaboration Showcase
20 Feb 2026 18:00h - 19:00h
Science AI & Innovation_ India–Japan Collaboration Showcase
Summary
The panel explored how “AI for Good” can democratize access to health and social services by simplifying and speeding up delivery mechanisms [1-3]. Kritika Sangani explained that Indus Action partners with governments to embed technology, policy redesign, and capacity-building into existing social-protection systems rather than creating parallel structures [14-19], aiming to cut the current ten-step entitlement process down to a single-touch interaction [24-26]. Using the Right-to-Education Act, Indus Action built an open-source digital public good that replaced a manual lottery with a digital one, scaling applications from 196 admissions to over 900,000 children across 18 states [70-82][84-89]; to improve targeting of the most vulnerable, the organization is piloting a multilingual WhatsApp chatbot that automates initial applications and supports frontline workers, thereby leveraging AI for discovery and outreach [91-98]. Himanshu from Atli Innovation Mission described the creation of State Innovation Missions to bridge regional disparities, citing examples such as a hackathon to map iron-contaminated water using AI and a dashboard to connect bamboo producers with global markets [122-164][165-176], and also noted broader AI-enabled public-service use cases including traffic-flow optimization, satellite-based lake monitoring, and sensor-driven leak detection in water pipelines [209-224]. Rajesh Babu illustrated AI’s potential in healthcare by developing an AI-driven briefing tool for pharma sales reps and an organ-matching system that evaluates numerous biological parameters to improve transplant outcomes [258-271][300-305]; he argued that AI agents can act as personal assistants for clinicians and patients, streamlining information flow and reducing waiting times for specialist care [280-295]. When asked about the risk of AI widening digital divides, participants emphasized that AI can be embedded with equity algorithms-such as gender-balanced lottery allocations-and that human-in-the-loop designs ensure frontline workers retain control [380-388][393-398]; Himanshu added that widespread smartphone penetration and multilingual language models help flatten urban-rural and linguistic gaps, though mentorship and venture capital still lag in less-developed states [400-417]. The discussion concluded that AI, when integrated as a digital public good and coupled with proactive equity safeguards, can accelerate social impact across education, health, and livelihoods while mitigating exclusionary effects [109-116][423-424].
Keypoints
Major discussion points
– AI as a tool to democratise access to social protection and welfare - Kritika explains that “AI for good… is about enabling equitable access for vulnerable citizens to social protection” and describes how the Right-to-Education (RTE) digital public good reduced a 10-step process to a “single-touch” digital lottery, now being used in 18 states [23-26][70-82]. She also notes the use of a multilingual WhatsApp chatbot to target the most vulnerable families [95-98][91-93].
– Government-led innovation missions to scale AI and reduce regional disparities - Himanshu outlines the role of the Atli Innovation Mission (AIM) as a “federal body that manages innovation for the country” and its focus on “setting up a State Innovation Mission” to bring AI to under-served eastern and northeastern states, including hackathons on water-quality data and bamboo-market dashboards [35-42][122-131][140-148][155-162].
– Private-sector/start-up perspective on AI’s impact and sector focus - Kavikrut and later participants stress that AI is “the strongest tool that startups have ever had access to” and argue that founders should channel it into high-impact sectors such as healthcare and education rather than chase VC money [55-64][61-64][358-367][377-382].
– Concrete AI applications across domains - Examples shared include:
• A multilingual chatbot for RTE admissions [95-98];
• An AI-driven “morning presidential briefing” for pharma sales reps that pulls past CRM conversations [261-267];
• AI-assisted organ-matching using multi-parameter biological data [300-307];
• AI-enabled water-pipeline leak detection and satellite-imagery monitoring [215-224].
– Ensuring equity and avoiding a digital divide - Kritika proposes embedding equity algorithms (e.g., gender-balanced lottery, SES targeting) and keeping “human-in-the-loop” workers like Anganwadi staff [380-388]; Himanshu adds that AI can bridge language gaps across India’s 22 scheduled languages but warns that mentorship and VC access must also be spread evenly [400-410]; Rajesh argues AI is a “flattening” force that can reduce digital inequality [393-398].
Overall purpose / goal of the discussion
The panel was convened to explore how artificial intelligence can be harnessed for “good”-specifically, to democratise access to public services, accelerate social-impact innovation, and create scalable, equitable solutions for India’s vulnerable populations. Participants shared experiences from government, non-profit, and private-sector perspectives, highlighted concrete use-cases, and debated how to operationalise AI responsibly at scale.
Overall tone and its evolution
– Opening (0-5 min): Optimistic and collaborative, with participants expressing enthusiasm about AI’s potential to increase access and speed [1-4][27-30].
– Middle (5-20 min): Becomes more informative and technical, detailing specific programmes, regional disparities, and concrete AI pilots [35-42][70-82][122-148].
– Later (20-35 min): Shifts to a reflective and cautionary tone, acknowledging challenges such as digital/linguistic divides, the need for equity safeguards, and the risk of “race to the bottom” [380-388][400-410].
– Closing (35-45 min): Returns to a hopeful, call-to-action stance, emphasizing collective responsibility to build AI for good and summarising key take-aways [418-424].
Overall, the conversation moves from enthusiastic vision-casting to grounded examples, then to critical reflection, and finally to a unifying, forward-looking conclusion.
Speakers
– Kavikrut – Moderator/Host, associated with T-Hub (startup incubator/accelerator) [S12]
– Kritika Sangani – Chief of Staff at Indus Action; development sector professional, former Teach for India fellow; focuses on social protection and AI for Good [S1]
– Himanshu AIM – Representative of Atli Innovation Mission (federal body under NITIO, public-policy think-tank of the Government of India); leads programs such as “Setting Up a State Innovation Mission” [S5]
– Rajesh Babu – Speaker on AI applications in pharma/healthcare (AI-enabled personal agents for medical reps, organ-matching AI); role/title not specified in external sources
– Audience Member – Unnamed audience participant who asked a question about medical breakthroughs and awareness; no role/title provided
– Audience Member 2 – Unnamed audience participant who asked about sectors needing more startups; no role/title provided
– Yashi Audience Member 3 – Unnamed audience participant who asked about preventing AI-driven digital divides; no role/title provided
Additional speakers:
– None identified beyond the list above.
The panel opened with moderator Kavikrut stating that artificial intelligence (AI) is set to “create access” and “democratise access to healthcare” by lowering both price and availability barriers [1-2]. He reinforced this optimism by highlighting access and speed as the twin themes of AI for Good [27-29] and warned that AI could trigger either a “race to the top” or a “race to the bottom”, insisting that the former can only be achieved by focusing on impact [63-64]. After the audience question, Kavikrut later urged founders to channel their talent into high-impact domains such as healthcare and education [377-382].
Kritika Sangani (Indus Action) situated her work within this vision. The organisation positions the government as the “protagonist” and itself as an “enabler”, embedding technology, policy redesign and capacity-building into existing social-protection systems rather than creating parallel structures [14-19]. Working with about 18 state governments and national ministries such as Labour, Social Justice and Employment, Indus Action seeks to make welfare delivery “equitable” for vulnerable citizens [20-23]. For Indus Action, AI for Good means collapsing the current ten-step entitlement process into a single-touch interaction [24-26].
A concrete illustration of this ambition is the Right-to-Education (RTE) digital public good. The programme draws on Section 12.1c of the RTE Act, which reserves 25 % of seats in private unaided schools for children from economically weaker sections [72-75]. Initial on-ground campaigns in Delhi generated 19 000 applications in 2013-14, yet only 196 students received admission calls, exposing a severe bottleneck [79-80]. In response, Indus Action co-developed the “RTE MIS”, an open-source, modular digital lottery that replaces the physical draw and reduces the transaction to a single digital step [81-82]. This solution has since been adopted, in various forms, across 18 states [82] and has enabled roughly 900 000 children to enrol in schools, up from a few 10 000 a decade earlier [84-89].
To further improve reach, Indus Action is piloting a multilingual WhatsApp chatbot that serves as the first interface for parents applying under RTE [95-98]. The bot not only automates the initial application but also supports frontline workers by lightening their workload and by using AI-driven targeting to identify the most vulnerable families [91-93][332-334]. Kritika emphasised that the same data ecosystems can be “flipped” so that the state discovers eligible citizens, rather than requiring citizens to discover schemes themselves [330-334]. An equity algorithm embedded in the lottery ensures gender-balanced allocations and representation of children with special needs or from socio-economically weaker groups [350-355][380-384].
Himanshu (Atal Innovation Mission, AIM) broadened the perspective to the national level. AIM, housed under the National Institution for Transforming India (NITI Aayog) and created in 2016 as a “brainchild of the Prime Minister”, coordinates innovation across ministries, from schools to high-tech startups [35-42][37-40]. A key current priority is the establishment of State Innovation Missions to bridge the stark technology gap between the more advanced western and southern states and the lagging eastern, northern and northeastern regions [124-130]. The first such mission is slated to launch in an unnamed northeastern state, aiming to create a peer-to-peer learning network that will allow less-resourced states to benefit from successful models elsewhere [140-148][233-235].
Within these state-level pilots, concrete AI-enabled projects were showcased. One hackathon will use AI to map iron-contaminated water at sub-district granularity, providing data for low-cost diagnostic kits [155-162]. Another initiative builds a dashboard to connect bamboo producers in a high-quality-producing state with global markets, addressing price-discovery challenges [165-174]. Additional public-service use cases mentioned include AI-optimised traffic-light cameras to reduce fuel consumption [211-213], satellite-imagery monitoring of drying lakes [215-218], and sensor-driven leak detection in water pipelines [220-224]. These examples illustrate how frontier technologies can be layered onto existing government data to generate scalable, locally relevant solutions.
Returning to the startup ecosystem, Kavikrut described AI as a “super-charged tool” for startups, the most powerful capability they have ever had access to [55-62]. He noted that AI is also reshaping the gig-economy, citing platforms such as Swiggy and Zomato that use AI for demand-supply matching [55-62]. Kavikrut also referenced a joint Swissnext-India startup project that is building an AI-driven test to predict cancer risk from DNA data, currently in validation [340-345].
Rajesh Babu illustrated how AI can augment professional workflows and clinical outcomes. He first cited an earlier, less successful attempt using Amazon Alexa Lex and Polly-the “Agilesium” prototype for a pharma-rep briefing tool-that was shelved due to poor language handling [260-266]. His team later built an AI-driven “morning presidential briefing” that aggregates a pharma sales rep’s calendar, CRM history and recent doctor interactions into a concise voice memo, dramatically improving information availability at the point of care [261-267]. More ambitiously, a collaboration with the Scripps Institute is developing an AI model that evaluates dozens of biological and physiological parameters to identify the optimal liver-transplant donor-recipient match, a task that traditional algorithms cannot handle [300-307]. Rajesh argued that, because smartphones are ubiquitous, AI will flatten rather than deepen digital divides, acting as a universal equaliser [393-398].
While there was broad agreement that AI can democratise access, the panel diverged on how to safeguard equity. Kritika stressed the need for proactive equity algorithms (gender balance, socio-economic targeting) and for keeping frontline workers “in the loop” so that human judgement complements automated decisions [380-388]. Himanshu added that India’s linguistic diversity-22 scheduled languages and myriad dialects-constitutes a separate divide that must be addressed through multilingual AI models [403-410]. By contrast, Rajesh suggested that the sheer penetration of smartphones already ensures equal access, implying that additional safeguards might be unnecessary [393-398].
Infrastructure was identified as a critical enabler of these ambitions. Kavikrut noted that India’s average 5G mobile data consumption of 22 GB per month provides a robust backbone for nationwide AI deployment [206-208]. Coupled with near-universal smartphone ownership, this connectivity supports both the United Entitlements Interface (UEI) vision-a UPI-like single portal where citizens can check eligibility and claim any constitutional right in one step [112-116]-and the multilingual chatbot and other frontline tools.
In the closing remarks, the panel reaffirmed that AI for Good must be built as a digital public good, embedded within existing systems, and equipped with equity-by-design safeguards [109-116][380-384]. Consensus emerged around three pillars: (1) simplifying entitlement delivery through single-touch AI interfaces, (2) leveraging state-level innovation missions and robust connectivity to level regional disparities, and (3) ensuring inclusive outcomes via equity algorithms and human-in-the-loop designs. Remaining challenges include scaling awareness of welfare schemes among the rural poor, finalising the governance model for the UEI, and establishing clear pathways for converting grassroots innovations into viable startups [326-334][378-382][393-398]. The discussion concluded with a collective call to “build AI for good” and to prevent a “race to the bottom” by prioritising impact, equity and collaboration [418-424].
Yes, you would think it will create access, it will democratize access to healthcare. Yes. Both in terms of price, in terms of availability. Yes. That’s great. Kritika, over to you. Tell us about yourself, the organization, and your take on AI for Good.
Sure. Thanks. Thanks, Kavi. Really happy to be here and privileged to share the stage with each of you. I’m Kritika Sangani, and I work as Chief of Staff at Indus Action. I’ve been in the development sector for about 10 years, started in the corporate sector. Serving investment banks in another lifetime. I’m also a Teach for India fellow. So I joined Teach for India and decided to not look back and continue with this sector. And, in fact, have spent, have been associated with Indus Action for almost 10 years now. Indus Action, as Kavi also alluded to, we work with governments on making social protection accessible to vulnerable citizens. The end goal is that we ensure that. welfare is delivered to them during critical life moments for the household such that they are able to actually tide over moments of crisis and make the most of moments of opportunities like education or healthcare to ensure that they’re trying to codify pathways out of poverty for themselves and their families.
The government is the protagonist in our work because, of course, they are the biggest implementers of social protection. We are the enablers. We come in with tech solutions, policy redesign solutions, capacity building solutions as a team. Our model is to embed these solutions within existing systems instead of creating parallel systems, and that is what we are about. We work with about 18 state governments, actively working with national ministries like Labor and Ministry of Social Justice and Employment. We work with the Department of Health and Human Services, And yeah, very excited to be here. I think when I listened, I heard about the prompt, Kavi, as what AI for good, what does it mean to us really?
It actually is about enabling equitable access for vulnerable citizens to social protection. Right now, they take about 10 steps, 10 burdensome steps to access a single entitlement. How do we bring that down to a single touch process? That is what AI for good stands for us.
Great. I heard access. You’re talking about simplicity and speed. I’m really curious to find out more themes as we keep talking. Thank you, Kritika. Himanshuji, over to you.
Yeah. Thanks. Thanks, Kavi. And thanks to, I think, the government of India for managing such a humongous event. I know as a matter of fact because we’ve been part of the planning team, almost a year’s preparation has gone into it. So kudos to everybody who’s sort of pitching in their own respective roles. So I’m Imanshu. I come from an organization called Atli Innovation Mission, which is the federal body that manages innovation for the country. We are housed under NITIO, which is the public policy think tank of the government of India. Atli Innovation Mission was the brainchild of the current Prime Minister, Honorable Narendra Modi ji. 2016, when we felt as a country that you need a body for innovation that cuts across all ministries, all life cycles.
So essentially it works from school to startups. And as a startup, even the high -tech startups like space. So we’re going to celebrate 10 years next week. The new title that we have sort of decided for ourselves is School to Space. Right. covering almost everything. I’ll not drain by talking too much about the programs that we do. Maybe we’ll talk it in a bit. I think for me and for the organization as a whole, because you’re also housed within a government institution, so social good means that whatever AI is trying to enable, is it having some impact at the grassroots level? Whether we work with incubators like T -Hub or startups directly or the state government so I also lead a program called Setting Up a State Innovation Mission, for example, where the idea itself is that how do we move beyond the current level of innovation that is happening?
How do we bring every part of the ecosystem together and put a layer of AI? Not only AI, I would say all frontier technologies. For example, when you talk to mature states in the southern and the western part, there’s this huge conversation on quantum. There’s huge conversation around even AI for health, for example. There are conversations where Can we enable AI to improve the public service delivery, for example, right? Where what we say we are not building unicorns, but we are building or we are saying not unicorn in terms of valuation, but in terms of social capital, right? Where we say that can it start impacting a billion lives, right? Yeah. So I’ll pause here.
No, this is great, Himanshuji. Thank you. We heard access. We heard simplicity and speed. You talked about scale and infrastructure because that’s the mandate with which organizations like AIM come from. I’ll talk about one bit on what we see at T -Hub. We have understood this very clearly that as T -Hub, as organizations, even like AIM, the goal is not for startups to be verticals or horizontals. Startups, MSMEs, nonprofits, I think we’re all here as vehicles to implement all this. Of economic growth. If you talk to startups, they don’t talk about policy. they don’t talk about infrastructure, they’re talking about building when they’re talking about building they just want to solve problems, they want to create value and I think AI has become a supercharged tool for startups to do that so the point I was trying to bring out at least from our experience with startups is that at least in the last 15 years, we see it all as a wave but it is probably the strongest tool that startups have ever had access to and a lot of social impact startups, even if you look at the Swiggy’s of the world and Zomato’s and if you take a step back and if you look at what they’re talking about on social media in the last let’s say a few months, both positively and negatively, you will see the major uproar is about gig economy and I was reading an article on The Economist which said that we are one of the only countries or economies in the world where the gig economy has become a true form of employment right so people are talking about minimum wages they are talking about you know labor treatment they are talking about human rights talking about all of that now who would have thought that a food delivery company right food delivery multiple food delivery and grocery delivery companies will actually create an organized labor market in our country so that’s the lens that we take for startups I will hand it back to you I have a question for some of you based on what you said but I will also drop in one theme here now when I said AI for good I think it obviously means AI for good impact you know for economic growth but it also means AI for good which is the other way to you know interpret the English of saying that it’s here to stay it’s here forever and it is our job to now figuring out what to do with it right and I’ll quote I think Nandan Nilkeniji yesterday said something about you know either it is a race to the top or race to the bottom with AI.
And I think the only way to go to the top is to focus on impact. So, Kritika, I want to pick this up with you. We know that intersection, when you talked about equitable rights, you talked about social protection, you talked about welfare and access to welfare and reducing 10 steps. Tell us a little bit of the work that you’ve done in RTE. I don’t think the audience is aware. Talk about admissions. And then tell us, where do you see the biggest opportunity for reducing those steps in that example with AI?
Absolutely. Thanks, Kavi, for setting that up. So, intersection started with the Right to Education Act. There is a specific clause under it. It’s called Section 12 .1c. Most of you would have watched this movie called Hindi Medium or English Medium, both versions. So, it essentially mandates that 25 % seats in private, unedited schools be reserved for children from economically weaker sections and disadvantaged groups. Now, we picked this sliver of a large constitutional mandate, which is Section 12 .1c, and with a fundamental belief in the power of choice for parents from vulnerable sections of the society to put their children in a school of their choice and not whether it’s public or private. And with that spirit, which is also part of the letter and spirit of Section 12 .1c, we started working on this particular right.
So we started our work in Delhi. I think we started with running on -ground awareness campaigns. And very early we realized that we were able to mobilize about 19 ,000 applications in 2013 -2014 in Delhi, of which only 196 students got a call for admissions. And this is a constitutional right. This is their right to get into the school. that was a huge eye opener for us and we realized that only by working with citizens and parents we are not going to be able to solve this problem because I think the government system also needs streamlining and support and they were really trying to also there was willingness to execute but the government is also really constrained for resources tech capacity and so on so we went there and fortunately the government showed willingness to actually experiment with a full online approach Rajasthan government had already done it so some of the peer effect also worked in our favor and that’s when we introduced what we call the RTE MIS that was our first solution it’s now evolved into an education digital public good it’s an open source modular product that we’ve launched for any government to adopt but what that did was what parents were going through was they had to it was it was this particular act it actually works on a lottery mechanism for selection So the draw of lots the school has to do, and the parent has to go to 10 schools that they have applied to, to actually see which child has, if the child has been selected in A, B, C, or D.
We cut that entire physical transaction and we got a digital lottery integrated, which is actually our secret sauce. Yeah. And that particular module has now was now adopted in some shape or form across 18 states that we’ve sort of worked with.
How many applications in total and children in schools now?
Children in schools now are about 900 ,000, 9 ,000, 9 ,000 children from one 96, 10 years ago, 10 years ago. And states, we actually when we started, we made an exit from our first state. Which is Delhi and then Uttarakhand. Uttarakhand adopted our end to end system. We took about seven to 10 years. Now we’ve shortened that entire cycle to three
And tell me, Kritika, to prevent, what is AI giving you as a tool to simplify and scale what you just described?
I think what we’ve actually now moved it to the next level by saying we are now focusing on who are these children who are applying to this right? Are they the most vulnerable? So it’s the challenge of discovery that we are using AI to for improved targeting for the state.
And how are you using targeting, AI in targeting?
So AI in targeting, what we are currently experimenting with is we have a WhatsApp chatbot. It’s a multilingual chatbot, which serves as a first interface for the students or the parents to apply to. And also using frontline, building frontline worker capacity. What that does is it actually reduces the load of the overburdened frontline worker with respect to reaching out to these students as well. and also being able to target the most vulnerable with respect to just having that multilingual advantage, having a physical touch point if I’m confused to actually reach out to somebody to navigate the system and then apply to the…
So the path from 196 to 9 lakhs was about 10 years, but I’m understanding that from 900 ,000 to 9 million will be much shorter. It wouldn’t be 10 years.
Absolutely. So there are 20 lakh seats available under this particular right.
Every year?
Every year.
Annually 20 lakh children?
Annually there are 20 lakh children. Currently there is about 60 % coverage. When we started it was at the 30 % mark. So we definitely…
So AI will scale that. And I think you didn’t talk about this, but I know from our other conversations that while AI will help you drive this deeper and scale, what you’re also doing is horizontally the DPG, the Digital Public Good they have built in education, the stack that they have built is now being universally applied to other entitlements or constitutional rights. So we heard the example of education. Absolutely. The dream, I know it’s called project name is UEI. They want to build a UPI like so instead of United Payments Interface, Intersection wants to build a United Entitlements Interface. Think of a DigiLocker meets a DigiYatra, meets the constitution. You log in, you check your own eligibility and you can tap into a constitutional right and apply for it and actually get access to something that is already rightfully yours.
Absolutely. This is phenomenal. Thank you so much for that, Kritika. Himanshuji will take a step back. I love depth, especially when people who are in action talk about it. Tell me where are you seeing, pick a specific example and where are you seeing AI truly unfold the real impact of what the work that you guys do. I have the same question for Rajesh Babuji after this.
Okay. I’ll probably step maybe two steps even further back, right? When the cabinet approved our extension in 2024, one of the programs that was really high on priority was setting up a state innovation mission, right? We did a couple of workshops and we realized that there’s a huge disparity between the western and the southern parts at one end of the spectrum and the northeast, eastern and northern part of the country, right? So Telangana, Andhra, Karnataka, Maharashtra are already talking in a level which is more towards the evolved countries as a whole, right? Let’s not say that these are parts of India in terms of technology enablement and even understanding of how this can be leveraged by a startup.
And when you take a diametrically opposite view to even the startups in northeast and even eastern part of the country, right? They are really, really not even at the basic level, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right? Even in the eastern part of the country, right?
Even in the eastern part of the country, right? new to all of these states. So one of the first few state innovation missions that we plan to set up is in the northeast, right? And the eastern part of the country. The first launch will be next week. I don’t want to take the thunder away. I know what’s coming up. We’re talking to them about it. You’re launching a state innovation mission next week. I won’t name the state. Yes, correct. And some of the conversation that we’re having with the state, right? And it’s a very progressive, evolved state. Not too much in terms of manpower or resources because of the simple reason that all the good people migrate to Hyderabad and Bangalore of the country and don’t remain in that state, right?
But those who have remained back, right? They’re actually building something really incredible for the state itself, right? They’re trying to solve some grassroot problems. I’ll talk something that is more relevant. And this is a conversation that happened about two months back when almost everything was set and the idea was how do we sort of launch a hackathon which is a ministry -backed hackathon. Now the state of, the state, let me not name the state, the water in the state has high content of iron, right? The idea was can we convert this into a hackathon and say that leverage AI, right, to first identify the content of water in different parts of even the district. So it’s not talking at the state level, going one level below district and even one level below district to a sub -district level.
Identify what is the different level of iron or different types of iron and then build a hackathon where a startup can leverage this data that has got collected, right? And build solutions two ways. One, a low cost diagnostic, right? Can we build a 100 rupee kit that can sort of diagnose this water, right? And second, maybe build a low cost solution to solve this particular problem. Other conversation that started happening. Now the state has a lot of bamboo, right? In fact, any of you. You would have been to the newer terminal at Bangalore airport. If you see all those beautiful bamboo things happening, it’s come from the state, right? The quality of bamboo is better than China and Vietnam, but they only get about one -tenth or one -twentieth the price, right?
One, because they don’t have access to market. Two, they are not able to even identify what bamboo will sell where at what price, right? Can we link it to the global market, identify who needs what? So they are trying to create a dashboard for it, right? And that thought came because when they realized that people have appreciated the bamboo that has gone into the terminal 2 of Bangalore airport, there could be a lot of use cases, right? Third element, and this was fascinating, right? And this is something which is sort of not something to be proud of in government. We have a lot of data, right? Everybody knows we have a lot of data in the government, right?
You are just trying to build a small funnel, right? That what’s the total? What’s the total number of innovations? who convert to a startup and create the jobs, right? Now, from startup to job is a slightly easier sort of, maybe multiplied by 10, 15, or even you have the data for these 100, 200 startups that are there in this district -registered DPIT startup. The first thought that came to my mind while I was talking to the secretary there, he said, let’s look at innovations that are happening in the state. Now, imagine the state has 120, 130 startups, but has 1 ,100 documented innovations which are grassroot innovations, right? And these are validated innovations because the overall innovation is 3 ,000, which is better than a lot of states when we took it in terms of per capita than even Karnataka and Maharashtra.
I come from a consulting background. The first thing I did was just divide it by the population, right? Look at a per capita thing. It’s twice the number for Maharashtra, right? And they are trying to solve problems which are very basic. For example, they are saying that, hey, if we mix this plant and this plant, that can lower your… blood sugar. I don’t know. Some of them may be patentable. Some of them may not be patentable. But the idea is that even if 10 % is good enough, right? So 10 % of that 1000 is about 100. It is just double the number of startups. Yeah. Where the innovation is already existing. Right? Somebody has to commercialize that innovation. Right? Right.
And this number sits on the National Innovation Foundation dashboard. Right? Which is, which was, and the state government had no clue about it. Yeah. That, oh, these many innovations. And then
And a lot of this you can now supercharge with AI. Yes. Which was slower before. And now we have multiple, let me just say, infrastructure layers, right? You have everything from 5G smartphone usage to the largest user base for chat GPT to, you know, I don’t know if you know. This number, but 22GB is the average usage of 5G internet. mobile data in India, 22 GB per month.
Correct. And the other thing that we are talking to this, how can we enable AI to improve the public service delivery when I was talking about that, right? And there are some broad use cases, right? Simple cameras on the traffic lights which monitors what is the flow of traffic at each hour or each minute, right? And LinkIt can be sort of automate that to reduce the consumption of the petrol, diesel, whatever it is, right? And then show some savings and fairly earn carbon credits. One very simple use case. Can we utilize, for example, satellite imagery to identify how lakes are drying up, right? Or what is the water? Just basic imagery. Everything is available. It’s not that we have to create newer thing.
And this is available with the government, right? One of the startups that we have funded at Artle Innovation Mission is a startup that has created small sensors that are embedded in the pipeline that tell you where, and measuring the flow of water at each level, maybe at every 10, 20, 30 feet, that where exactly is the leakage. So that people don’t have to go to that pipe and then see, oh, where is it going, let me just fix it up. You directly go to that. The next level that the startup is building is, can we also look at where the leakage is happening for the multiple times and look at what is the material, right? Is there some more stress at that particular point of the pipeline to ensure that next time that you’re building or repairing it, the leakage doesn’t come from this particular part of the pipeline.
No, this is great, Himanshuji. What I’m hearing from you is that, first of all, I can feel your sense of energy and excitement for the region. Upcoming state innovation mission launch as well as the beautiful vest you’re wearing today. This is not. This is not the state that. Yeah, we’re not giving any hints. And I was just going to say that what you’re saying is in like the pipe. example that you talked about is that there are existing innovations, there are existing latent energy that startups have, there are real problems that can be solved and AI is helping, you know, as an example, tie all of this together. It’s not just supercharging or fueling this, but it’s the glue that’s bringing all this together.
And the best part is that in truest sense, this is democratization because AI will enable that Assam and Manipur or Meghalaya or Sikkim are at the same level as maybe Karnataka because if you have that AI level on a data, right, and you want to take any data back decision for public good, right, it doesn’t stop Sikkim to take that decision which Karnataka can take today.
And the other very important thing that we are trying to build within this state innovation mission is to create a peer -to -peer learning network, right. There could be something fascinating that Sikkim can offer in terms of organic vegetation or agriculture which can be picked up from the other agrarian states. Yeah.
Yeah. This is great. I’ll, we’ll go over to Rajiv. Rajiv Babuji and then we’ll break for questions. You know, we’ve heard the building in public perspective from a non -profit as well as a federal organization. I want to see the private slash corporate, you know, and the work you guys do in the foundation perspective, Rajababuji. You already spoke about, you know, access. You spoke about availability in healthcare. Pick an example that you’re already building towards and tell us what are you most excited about, about how the products and services that you are building will truly unlock value for social good and economic growth.
Always. Yes. So I’ll first touch upon one point. See, there is a lot of concern and fear about AI coming and taking away the jobs, right? I think that’s partially true, but in the big picture, it is not true. See, what’s going to happen every time technology comes, for example, when humanity started, as a civilization we started, there was only one industry, which was agriculture. Yeah Pre -industrialization Exactly And then after that When industrialization happened More opened up It could have been seen Then with the power it came It saw very disruptive From horse power to steam power That could have been seen very disruptive To many people who Livelihood was depending on horse power Or bull power or whatever power it was Same way from steam When it moved to electricity It could have been very disruptive People whose business was all in steam Coal and all that stuff They would have thought Oh this is not necessarily a good thing It’s going to put all of us And even yesterday somebody was referring In early late 1800 They said we’ll close the patent office Because everything needs to be Innovated has been already innovated There is not much to do But then here we are Almost 130 -140 years later Of making the patent And we are still in the process Of making the patent And we are still in the process And we are still in the process Of making the patent And we are still in the process new things are unfolding.
So I think, you know, like in the, you know, many in the forum leaders spoke, A is like another energy, right, like electrification. It’s going to do a big transformative change, and there is not a thing it won’t touch. It will touch everything. And it’s going to make everything intelligent. And when everything becomes intelligent, largely, very good things will happen. Very good things, very positive things. It’s going to enhance the livelihood or quality of life for many people, and it’s going to create a lot more opportunities. With that being said, in the industry we are focused in, right, for example, I will tell one example which we tried, Agilesium. One of the things we wanted to do is that, so we work with pharma company, biotech company.
One of the problems we were trying to solve I will just give few problems Which we were trying to solve Was not unsolvable At that time because of the technology As recent as 4 -5 years
Tell us more about this example
For example in 2018 We invested in a project Our customers were Pharma companies It was reps So the reps basically are going And meeting the doctors They are supposed to know which doctor They are meeting What was the last time the conversation They had with the rep And doctor said I wanted to know About this medicine What kind of adverse impact it could have For the patients Side effects it may have if I give it to the patients Some technical details the doctor would have requested So they have to go back And maybe talk about it Give those details and all that stuff But between two appointments Sometimes it’s three months to six months So they don’t remember what happened So what we wanted is Last time conversation recorded Of course they have it in paper or sales force And all that stuff They may put it But then you are going today from the previous meeting You don’t remember that You are not accessible You are in the car driving and all that stuff So what we wanted to do is We wanted to do an AI Where in the phone app it’s implemented And you Previous day Based on it will look at your calendar It will look at who are all the doctors you are visiting And then it will go to the CRM What was the conversation Last two three conversations happened with the doctor You take that And what homework you did on that It will take all that Then it will take as a voice memo Morning 7am when you are going to meet the doctors You play that voice memo It says hey today you are meeting so and so doctors This is what was the conversation last three This is what you are supposed to tell It’s
almost like a morning presidential briefing Exactly,
morning presidential briefing And I got this idea based from that only Amazing then at that time The technology available for me Was Lex Lex is a Offering tool AWS had released from Alexa So Alexa, whatever they used to do Build Alexa They gave that tool as Alexa And then there was another technology called Poly, which also from AWS Technology So we took these two technologies For our customers from the AWS platform We were trying to build it It sucked At that time The experience was really bad Because They could not understand the keywords The medical keyword It would not understand People’s accent it would not understand And reading structured and unstructured data Because you have to read the unstructured data Structured data and all that stuff And you have to do it And you have to do it It was not so great Plunky now, so we have to shelve that project.
We invested almost a million and a half, two million dollars at that time, multiple customers of ours. And the experience when we took it to the rep, it was not so great. They did not If it’s not easy and usable, nobody adapts. It would not understand the question. Can you tell, repeat it again, repeat it again, three times, they would throw the phone. Right? It was like that. And when it got it also, it did not understand. It was bad. Because of the accent, various things, medicine, technical. Now, that has been implemented a year back with the AI super seamless.
And where do you see the impact of adoption of this technology?
Across the board, everywhere. Because everywhere if you think about it now, right, it can take every individual, every one of our individual, let’s say our calendar we can take. We can create a personal agent for it. and it will look at the task, it will look at our email, it will look at our calendar and it will tell this is what you are supposed to do. I
think a lot of people have begun to use that. Yes,
and then now not only self, an organization can create an agent for each and every buddy. Instead of manager going and telling each and every person, hey, you are supposed to complete this, we can have their personal agent who is their agent buddy can tell them, which is a little bit private and more comfortable to hear. And
I think in healthcare and the service that you are in, a continuous flow of information, high quality information, makes a huge difference in availability of the work that you do. Exactly,
this is simple as accessibility information like you said. And scale. But then I will take another complex situation where we are working with another research institute in San Diego. Yeah. Where… every individual, every one of our individual, let’s say our calendar we can take, we can create a personal agent for us. And it will look at the task, it will look at our email, it will look at our calendar, and it will tell, this is what you are supposed to do. I think a lot of people have begun to use that.
Yes.
And then, now, not only self, an organization can create an agent for each and everybody. Instead of manager going and telling each and every person, hey, you are supposed to complete this, we can have their personal agent, who is their agent buddy, can tell them, which is a little bit private and more comfortable to hear.
And I think in healthcare and the service that you are in, a continuous flow of information, high quality information, makes a huge difference in availability of the work that you do.
Exactly.
Which is access to medicine. This is simple as accessibility, information like you said. And scale.
But then, I will take another complex situation where we are working with another research institute in San Diego.
Yeah.
Where basically, this is in the liver transplant. Where the patient basically is waiting there Who is in the waiting list For a longer time The donor comes in, they would be matched Now what we are doing is That is not the way Because you know in the liver transplant Any organ transplant Absorption of that organ is very difficult Because it is seen as a foreign body
Yes, the foreign body
And the body will Try to reject it Making it accept is a very very big problem So then You need to look at the biologics Of both the patients You see who is most conducive From a biological point of view To receive this organ So there are parameters, biological parameters And physiological parameters that now you can do a match And there are too many parameters to do In a simple Old algorithm It is very difficult Now what we do, and then there is multiple research papers
How far along are you on this matching For organs
Now we are helping Institute One of the top institute in San Diego Scripps And lot of Nobel laureates from medical are there And we are helping One of the researchers actually from India It’s published so I can quote His name is Sunil Kurian He has published and he has implemented this with this and with the AI Now he has told based on the patients, donor and the patient The AI can tell who is the best recipient for this match And their livelihood, the organ, after that they will have a better living condition It can predict based on various parameters which is not very easy to do So from a simple use case as a
No it’s a phenomenal To very deep science You went from morning presidential briefings to almost Tinder for organs in one shot No this is great I want to ensure Rajeshji we have enough time for questions We have a small audience here I wish we could take live questions online too But we don’t have that facility yet But any questions here, small audience Any questions that we have Very happy to take them You can point it to a certain panelist if you like Yes please we will pass the microphones right behind you please tell us your name and you can point a question to any of us thank you very much
I have two questions one is for Rajesh ji and one is for Kritika ji so Rajesh ji what medical breakthrough do you believe will come up in the market in next 3 -4 years after this integration of medical science and AI is my question crystal ball question on healthcare what is the breakthrough on the healthcare medical side that you see it’s a crystal ball question that might happen in AI that you are excited about second question for Kritika your initiative is very impressive what I want to know is the average poor person somewhere in the rural area is he aware of your initiatives or if not then in what time frame do you believe that you will be able to reach that awareness question
I will add a flavor to Raj’s question of saying can that awareness be unlocked for further media We will first start with Rajababuji Please go ahead
Thank you, very good question, appreciated See, I think definitely Lots Lots, right, it is going to be very Transformative, it is going to be Both in the healthcare side First the healthcare, right, I think the Primary healthcare Lot of it will move to AI, right, I think Through our AI Basically doctors will Start creating the doctor agent And first your Questions, your personal doctor agent will Address it, basic, and then Based on that, your personal doctor agent Your doctor’s agent Will now next contact the doctor If there is a need
I have a feeling that the doctor’s agent will talk to your patient’s agent
Yes, yes
Before it comes to the two humans
Exactly
So what is the breakthrough for you in that
See, that Basically the access, no waiting Especially, you know, in the western countries, the waiting, some of these specialists is unbelievable, we would not have heard in India, they wait for six months to one year.
Right, that’s the waiting time for surgeries too.
Yes, so I think that will definitely transform and then also you said like, talked about one of the things, sir talked about the bamboo situation, why not a global healthcare? Like why can’t there be a marketplace immediately, they can reach and there is a doctor, of course already that is happening but that can happen in a much much greater level because these agents of AI could be sitting at hospital level, doctor level, patient level, as a patient, I may not be AI savvy to reach various things but my AI agent avatar which I have created, if we can create in some easy way that would reach it and definitely address it.
No, I think you brought up a very important point and we will go to Kritika right after is that, what you have said is that in highly skilled professions, where human capacity is at a very significant scarcity that capacity can be unlocked to a further level simply because agentic will come in. No that’s great I hope Yuvrajji the first question is answered I just add on to what you are saying right I think we were doing a joint program with Swissnext which is the Swiss arm of the embassy of Switzerland here and this was like a exchange kind of a program right and one of the Indian startups that actually spent a week in Switzerland what they are trying to create is look at all the profile of cancer patients right and start mapping that to the DNA and start predicting what is the probability right so they have created a test it seems still at a validation state what is the probability of you getting cancer looking at all the data patterns that they have in the database right and they feel that this will sort of enable a much faster cure for cancer because you’ll be able to predict it much better.
Yeah. Krithika, over to you. I’ll expand Ugra’s question again. I’ll recap. I think he’s asked something very interesting and important is can in a country of a billion plus people despite the availability of technology awareness of important, interesting impactful things continues to be a challenge. Given your experience at Indus Action, do you feel there is an awareness problem based on the work you’ve done with now 9 lakh students and can that awareness challenge be leapfrogged using AI? Over to you.
Absolutely. Absolutely. No, this is a really relevant question and something that we are very actively working on. So what we’re saying is we want to flip this question. Whoever the citizen is, citizen discover. That the citizen has to discover the scheme. Can we use AI? And tech to flip this and say can the state discover the citizen? and building in those layers of AI ML into existing data which is exhaustive and it is with the state. To say that what if I layer your VBG Ramji or erstwhile M .G. Narega data with PDS repository with say aspirational district data and say in this district almost 95 % citizens are eligible for this particular entitlement. And this currently happens for demographic longitudinal research.
Correct. Absolutely. It is not used for actual access. Absolutely. And actually there is an organization. So Educate Girls has a model wherein they have been able to use ML to be able to identify and pinpoint to say these are the districts which need an intervention. These are the districts where we have the most number of out of school girls. So awareness will flip to actual targeted outreach. That is exactly what we are now trying to attempt with the state. The problems of validation, verification of citizens, are they eligible, not. Can we use? Yes. Can we use AI? Yes. ML, tech to be able to flip this and say, let tech enable the stage to discover the fascinating.
And that will also reduce the 10 steps in one, in one single shot.
What an answer. Love this. I think we can take one more question. We have a few minutes left. We have to, we’ll take to keep your question short and pointed to, you know, uh, whoever you want to ask, please go ahead.
Thank you for a valuable session. Uh, I just wanted to ask that, uh, there are a lot of startups in India, but in which sector you think there should be more startups, but, uh, but aren’t currently.
Great question. Great question. Do you want to answer it? You and I can pick this up. You go first and I’ll follow which sector lose largely defined, loosely defined. Do you think there’s a huge volume opportunity for founders?
Okay. Uh, I think that it is a two level. One is if I define sector by, uh, what we traditionally call a sector, let’s say ed tech or a health tech, uh, or the second would be. Is it like a frontier AI, non AI kind of a thing, right? Uh, So, if you look at the current sectors, right, what most startups are doing, that’s my personal belief, not the organizational belief, right? What most startups are doing is they’re trying to, or they were trying to chase venture capital money. So, if I felt that EdTech is hot today, let me go EdTech. Today, that same thing is for deep tech, right? Just enable some element of AI or some element of deep tech and say, I am a deep tech startup, right?
The point is not that. So, I think one, founders should look at some problem to solve. I’m not saying a social problem. Some problem, some gap that they feel. With the current ecosystem or the current startups are not fulfilling. Or create your own niche. That’s one. Second, what some of our programs also do, and I did talk briefly about it, is the social kind of a thing, right? Do we create solutions that can sort of enable or solve for problems in India? Because a lot of problems that we solve for in India or in Indian context. Are replicable to the global south. Because similar per capita GDP, similar environment, similar diversity, similar price point.
you want to know that is great very very interesting perspective it’s a fantastic question i’ll just answer it with one simple perspective foundership is a long journey even if you look at the current ipos they were all been building for 10 -15 years ai will not shorten that journey but will make impact faster i believe so i think given the world that we are in right now and this massive phenomenal super tool that we are getting i think founders should build things that take time because that can be accelerated now these are essential problems if i have to pick sectors i would say healthcare and education and not because we have them here but i think even if 10 percent of the talent that’s currently in fintech or is currently in consumer tech i think if they move to healthcare it could be engineers founders investors it will fundamentally change this you know what this country can do and needs and it’s the same for education i don’t want to say unemployment but we have a massive massive young talent pool the largest in the world if we can change you know as founders and as startups we can build more tools for education i think we will unlock a superpower you know for our country so i think that’s where founders should go thank you one last question you had one to ask us
hello i’m yashi i wanted to ask one question anybody from the panelists can answer my question is how can we ensure that ai systems deployed for public welfare do not deepen digital divides especially for rural and marginalized communities
what an amazing question ai can deepen divides i will use the i’ll take the generosity to not just say digital it could be economic social uh you know cultural there are many divides that exist in the country how can it ensure uh kritika do you want to take a shot at it given that you’re in the welfare space i’ll ask maybe all of us can give a quick one line 10 second rapid fire answer on what can we do ahead proactively for the reducing a potential digital divide thanks to ai
yeah i think uh two perspectives uh One is in our solutioning, when we, and I’m going to take a live example, when we actually embed, say, a lottery algorithm within a state, what we tend to do is we also have this equity algorithm to say that the gender ratios are going to be balanced. It’s going to be the girl and the boy are going to have a 50 -50 application rate there, right? Or to say that this is going to be the percentage of, you know, children with special needs or children from, say, socioeconomically weaker backgrounds or like SESD, OBC. So one is, like, how do we proactively embed these algorithms to be able to address what you’ve just suggested, which is also a problem that we actively work on.
I think other is I definitely feel. I feel that we cannot discard the human in the loop. I feel like AI has to make. their jobs easier, give them my Anganwadi worker or my Asha worker. They need to have access just the way Rajbabuji was referring. They have information which is so simple and so easy to disseminate that they don’t have to spend hours and hours parsing through these complex eligibility criteria for welfare schemes and so on. Those are the two perspectives that I have.
Fantastic. You are basically saying build a pro -social equity bias into AI.
Absolutely.
Rajbabuji, do you have a quick rapid fire answer to this?
Thank you. I want to also answer the previous question because that was a very good question. I will quickly touch on that and come to this. See, I think like Kavi told, don’t chase the VC money. I mean, it should be always see the monetization should follow the value. What you should be focusing is Are you creating a value And what is happening is Many are chasing the monetization And missing the value Are you making a difference If you are making a difference And wherever it is, whatever small it is Like he said, health tech Especially a lot of potential is there We need to be making disruption there So focus on that But there are others also On digital divide and social equity On the digital divide I think AI with the phones already Everybody has a phone now Almost everybody is going to have a smart phone If not, already they have it Once they have a smart phone, they have AI So AI is not going to divide If anything, it’s flattening everybody out For example, the people with computer engineering They were on the pedestal Programmers Now it has brought innovation Not only for a programming person It has brought down to everything If anything, AI is the biggest equality It is not a divider
Fantastic Himanshuji, any view on that? What can we do proactively?
Yeah, I think I think sort of agree to what everybody has talked about. I think with AI and the smartphone and we are one of the largest consumers of Internet. Right. It has sort of democratized the divide between rural urban. I think there’s one more divide which we generally don’t talk about is a language divide. Right. There are 22 scheduled languages. No other country in the world has 22 scheduled languages. Right. And this is just scheduled languages. In a state like UP, it looks like Hindi, but you travel from the western part of the eastern part, the dialect changes. Right. So that’s another divide that probably is being sort of democratized by AI, for example. And everybody’s building solutions around it.
Right. All the LLM models that are being developed, both by the government and even some of these larger companies that want to look at our data. Right. So in the future, I think one thing that we need to be very, very cognizant about is which we I started with. Right. The divide between the western. And the southern states at one end of the. spectrum, eastern and the northern states at the other end of the spectrum, right? The challenge is that a lot of mentorship, VC money, all that has not reached it. So they are still trying to build those native solutions, which is not bad, right? But for them to really equalize with some of the other states, that’s where AI is going to enable.
And some of the state governments are building and all of us are playing our own little role in that.
This is great. You brought it all together Himanshree. We’ll wrap this panel now. Now, I think AI is here for good and it’s our duty to build AI for good. And what an interesting conversation. The divide is a matter of what we do. And it’s great to have all your perspective. Thank you everyone for joining us and for this conversation. Thank you so much. Thank you. Thank you. Thank you. This is the photo of the marathon. That’s why they take it at the start. You can check it out there. Please, thank you. Thank you. Thank you.
“Moderator Kavikrut stated that artificial intelligence (AI) will democratise access to healthcare by lowering both price and availability barriers.”
The knowledge base notes that AI is helping to democratise access to healthcare through early disease detection and wider availability of high-quality diagnostics [S16] and that leaders highlight AI’s role in making healthcare more accessible globally [S101].
“Kavikrut warned that AI could trigger either a “race to the top” or a “race to the bottom”, emphasizing the need to focus on impact to achieve the former.”
Sources discuss the concept of a “race to the bottom” in AI regulation and warn that it is progressing faster than a race to the top, underscoring the urgency to steer AI diffusion responsibly [S106] and note the dangers of a regulatory race to the bottom [S94].
“Kavikrut urged startup founders to channel their talent into high‑impact domains such as healthcare and education rather than chasing venture‑capital trends.”
The moderator explicitly encouraged founders to focus on high-impact sectors like health and education in his remarks [S1].
The panel exhibits strong convergence on the view that AI should be harnessed as an inclusive, democratizing technology—simplifying entitlement processes, embedding equity safeguards, and scaling impact through startups and state‑level missions. Participants across government, civil‑society and private sectors align on the need for robust connectivity, multilingual capability and human‑in‑the‑loop designs.
High consensus: most speakers echo each other’s core messages, indicating a shared strategic direction for AI‑for‑good initiatives in India. This broad agreement suggests that future policy and programme design can build on a common foundation of equity‑by‑design, single‑touch service delivery, and infrastructure‑driven deployment.
The panel largely agrees on AI’s transformative potential for democratising access to health, education and social services, and on the importance of digital infrastructure. The main points of contention revolve around how to safeguard equity—whether AI is automatically inclusive or requires explicit algorithmic and linguistic interventions—and the strategic direction for startups, i.e., sector‑specific focus versus problem‑first innovation.
Moderate. While there is consensus on the overarching goal of AI for good, the disagreements pertain to implementation pathways (equity design, language inclusion, sector prioritisation). These differences suggest that policy and program design will need to balance optimism about AI’s equalising power with concrete measures to address bias, language diversity, and sectoral strategy.
The discussion was steered by a series of concrete, ground‑level examples that transformed the abstract notion of ‘AI for Good’ into actionable pathways. Kritika’s focus on simplifying entitlement access and embedding AI in existing public systems introduced the central problem of bureaucratic friction. Himanshu’s emphasis on regional disparity and language barriers reframed AI as a tool for equity, prompting the panel to consider both geographic and linguistic divides. Kavikrut’s ethical framing (race to the top vs. bottom) and Rajesh’s vivid illustrations—from personal agents for pharma reps to AI‑driven organ matching—added depth and breadth, moving the conversation from policy to daily workflow to cutting‑edge science. Each of these pivotal comments sparked new sub‑topics, shifted perspectives, and deepened the analysis, ultimately shaping a dialogue that balanced visionary ambition with pragmatic, inclusive implementation.
Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.
Related event
