# Day 51 — YouTube 自动生成字幕（AUTO_GENERATED）

覆盖训练段 000–12 分钟。已去滚动字幕重叠。ASR 生成可能含识别错误。

==================================================

Hello and welcome. You're watching Global Lens with me Parikhit Lutra. These are the top global headlines uh we are tracking and the big focus today is the AI impact summit 2026 which is set to take place in New Delhi from the 16th of February to the 20th of February positioning India at the center of the global conversation in artificial intelligence. The 5-day summit is expected to bring together policy makers, technology leaders, researchers, industry voices from over 100 countries to discuss the next phase of AI development

and real world deployment. So what are Indian companies hoping to see from this platform? Faster pathways for real world AI deployment, better access to compute and data infrastructure, deeper global infrastructure, deeper global partnerships for sector specific innovation. Industry leaders expect stronger momentum in applied AI particularly across healthcare, particularly across healthcare, agriculture, public service delivery where outcomes are measurable and scalable. The summit is looking at healthcare as a priority sector for AIdriven transformation. In the context we put this uh spotlight on Nam AI, a Bengaluru based

healthcare company using AI powered thermal imaging for early breast cancer detection. Its non-invasive radiationfree screening non-invasive radiationfree screening technology is designed to be privacy preserving and affordable making it especially suited for large-scale rural outreach. Backed by 39 patents, Nurum AI has screened over 300,000 women across 22 countries and is working with hospitals, diagnostic chains and public health programs to expand access to early detection through AIEL diagnostics. diagnostics. Joining me now to discuss this further is Gita Mangjinad, founder Nuram AI. Uh Dr. Mangjinad, thank you very

much for being with us. Let me begin by asking you about uh your experience in scaling up AI in India and across the world. What is the milestone you seek to achieve in 2026? achieve in 2026? >> Uh first of all, thank you so much for uh you know uh featuring NAMI or also called NIAM.AI on this show. As uh you mentioned uh we have developed a breakthrough innovation in cancer screening particularly focusing on breast cancer screening uh where uh AI is the primary focus and

differentiator. We can we use 25 plus machine learning algorithms to convert temperature maps to a cancer health report to make an affordable accessible uh breast cancer screening for everyone and also in a privacy preserving manner and uh in a portable way so that anyone can take breast screening at their doorsteps. Um so far you know u we've learned a lot from the ground. First thing we have learned is India is a ground for innovation because we have lots of problems to solve and lot of constraints

uh you know while solving the problem. So uh we all know that uh necessity is the mother of innovation and so that's how we've developed this AI based breast cancer screening on the ground and what we found is that India requires a way of enabling this at a a very very affordable manner in order to reach the corners of uh the village. So in order to create that impact uh on on everybody and so we had to devise the solution in a way that uh we

have different product models. one AI model which is giving a lot of information for the doctors to take further uh diagnostic workup uh and another AI model which can pretty much work independently just in the hands of a healthcare worker an asha worker who may not have a lot of skills from a clinical perspective but she's the one who will communicate communicate to the end users so instant triaging for one side and detailed reporting on the other side there are several things we learned but happy

to share All right. I would also like to ask you what is the challenge that one faces in terms of scaling up uh AI applications like NIAM.AI like NIAM.AI uh with the government because clearly if you want to provide these applications at scale in different languages you also need the government as a stakeholder. So what is the challenge that companies like yours face? face? >> Yeah, you're absolutely right. uh especially in country like India it's very very important to partner with the government to reach to

a lot number of uh women and scale up uh so in this case I think one of the main challenge we face is that health is a state uh decision making uh you know entity right so basically uh health decisions are made at the state level and uh very little can be controlled at a central level so we need to go to every state convince the stakeholders including uh of starting from the health minister or ending at the health minister but all the people in the

chain and so it does take a lot of effort to convince them that AI can actually be much better um uh than not having a doctor at all right you know having an AI enabled tool for let's say a triaging um instant triaging to identify high-risisk women or high-risisk patients in the remote areas and bring them to a more detailed workup in in the uh urban areas where the infrastructure is available is definitely feasible. So we have to do a lot of feasibility studies, pilots almost

for every state and that's usually free of cost and it becomes a little bit uh harsh on a startup like us. So to everyone you have to prove that it works and after that we need to figure out what is the pathway to adoption. So we still sort of working on that level but uh we've had a fair amount of adoption uh in Punjab, Maharashtra, a little bit in Karnataka uh now in Hana and so on. >> Right. I would also like to ask you uh

what are some of the trends that you see across the world with new patents that are being filed uh in terms of AI for healthare users? What are going to be some of the breakthroughs that we should watch out for over the next two to three years? years? Yeah, it's a great question because whenever we talk about AI and future of AI and and where it goes, uh we seem to be focusing more on the infrastructure level like you know how can we make of course

infrastructure is important we have to make fast decisions we have to have sort of deep learned networks like LLMs and GPTs etc. those are important but those are not sufficient enough right so I think the patents that we need to do more and so far we have filed and granted 39 patents including 15 uh granted in the US so here we are focusing on what are those new biomarkers that we need to extract from this data that we already have so that we can personalize each

of these diagnosis screening treatment what say what what what have you uh depending on the application the application So how do we almost look at the DNA or radiomics radiomics extracted from AI from this data that's where I think we need to focus more patents on and that's where our initial patents have been and it has been uh you know able to trigger and fuel the breakthrough innovation that we have done but most of the focus is going towards infrastructure and yes you need the machines

we need the platform but the healthcare application uh is is extremely complex Because number of people who are healthy uh is much much more than number of people who are unhealthy unhealthy >> but unfortunately you have to find those unhealthy women or men as early as possible and as accurately as possible. So that is where the domain specific innovation is absolutely needed and new patents uh should be seen. >> All right, my final question. What is the kind of conversation you hope will take place on

healthcare at the uh AI impact summit in New Delhi and what is the one thing we need to do uh to provide an enabling framework to companies like yours? companies like yours? Um I am really really excited to be part of the U AI impact summit and I'm so happy that it's happening in India because India is going to define the future of AI which is impactful um and so so what I want to showcase there is definitely our solution. We have got invitation to put

our solution a demo uh case in multiple stalls and we will put in there and showcase to the rest of the world how we can really use AI to touch uh so many lives and save so many lives that we have already done. So and with this we want to take our make in India solution to several similar low and middle inome countries where you see a screening rate of 2% like 98% of women are detecting their cancer by hand. So I feel that this platform

will do two things. One, put India on the top of the AI impact uh country I would say leader. Uh secondly, it will enable several of our startups like ours to reach out to similar countries and expand our uh reach beyond India and and impact so many women around uh out there. >> All right. Thank you very much uh Miss Mangjunat for joining us with your experiences with NAM.ai AI and what lies ahead for AI in healthcare. We now shift our focus to Nasa AI, a

Mumbai based startup founded in 2023. The firm provides AI acceleration cloud infrastructure for enterprises in India. It offers platforms like velocis for GPU based training and inference emphasizing sovereign AI with openw weight models. It ensures data control, cost efficiency and compliance over global hyperscalers like AWS or Azure. Joining me now to discuss the future of AI in India is Nasa.ai's founder and CEO Shahhat Sani. Shahed let me start by asking you how does the union budget help companies in scaling up AI infrastructure. So Paris first

of all thank you for having me and u I think the tax holiday that the government announced in the union budget for global cloud firms to set up infrastructure in India helps a new cloud provider like us because we can now be in a position to cater to their workloads in India giving us the scale that we need. uh in addition to that uh India's uh the union government has also reiterated their commitment for uh budget for the AI mission. So I think both these announcements

actually help us because they allow us to invest in uh the AI compute network and storage not only for the Indian audience through the AI mission but also for the global cloud service providers. uh also how do you see the funding pipeline for AI companies in India? >> So Parish I think NSA may not be the right example. I think since I'm a second time entrepreneur and I had a very successful first company, I've not had it difficult. Uh but um and so we've been able

to raise money from marquee investors whether it was Nexus Matrix or Entity VC and we're about um hopefully next week we'll announce a larger funding round. But um I think um in general if for startups in the AI space if they've been able to demonstrate um you know um client success and if they've been able to demonstrate uh something that's uh unique uh you know most VC firms and most private equity firms are looking for for um companies startups in this space. So virtually every single

major VC firm wants a start is looking at startups with some proven track record or with some kind of customer validation in the AI space. So it's actually most of the funds are

# Day 51 — 火山引擎 录音文件识别（volc.seedasr.auc）

豆包大模型 ASR，异步接口。可能含识别错误，仅供听力对照。

==================================================

Hello and welcome.

You're watching Global Lens with me, Parikshit Lutra.

These are the top global headlines we are tracking.

And the big focus today is the AI Impact Summit 2026, which is set to take place in New Delhi from February 16th to February 20th, positioning India at the center of the global conversation.

And artificial intelligence.

The five day summit is expected to bring together policy makers, technology leaders, researchers, industry voices from over hundred countries to discuss the next phase of AI development and real world deployment.

So what are Indian companies hoping to see from this platform?

Faster pathways for real world AI deployment, better access to compute and data infrastructure, deeper global partnerships for sector specific innovation.

Industry leaders expect stronger momentum in applied AI, particularly across health care, agriculture, public service delivery where outcomes are measurable and scalable.

The summit is looking at healthcare as a priority sector for AI driven transformation.

In the context, we put this spotlight on Nirum AI a Bengaluru based healthcare company using AI powered Thermal imaging for early breast cancer detection.

Its non invasive radiation free screening technology is designed to be privacy preserving and affordable, making it especially suited for large scale rural outreach.

Backed by 39 patents, Nirum AI has screened over 300,000 women across 22 countries and is working with hospitals, diagnostic chains and public health programs to expand access to early detection through AI LED diagnostics.

Joining me now to discuss this further is Gita Manjunath, Founder Niram AI uh.

Doctor Manjunath, thank you very much for being with us.

Let me begin by asking you about, uh, your experience in scaling up Niram AI in India and across the world.

What is the milestone you seek to achieve in 2026?

Uh, first of all, thank you so much, uh, for, uh, you know, uh, featuring, uh, Niramai or also called Niram dot AI on this show.

As, uh, you mentioned, uh, we have developed a breakthrough innovation in cancer screening, particularly focusing on breast cancer screening, uh, where, uh, AI is the primary focus and differentiator.

We can, we use 25+ machine learning algorithms to convert temperature maps to a cancer health report to make an affordable, accessible breast cancer screening for everyone and also in a privacy preserving manner and in a portable way so that anyone can take breast screening at their doorsteps.

So far, you know, we've Learned a lot from the ground.

First thing we've Learned is India is a ground for innovation because we have lots of problems to solve and lot of constraints, you know, while solving the problem.

So we all know that necessity is the mother of innovation.

And so that's how we've developed this AI based breast cancer screening on the ground.

And what we found is that India requires a way of enabling this at a very, very affordable manner in order to reach the corners of the village.

So in order to create that impact on everybody.

And so we had to devise the solution in a way that we have different product models, one AI model, which is giving a lot of information for the doctors to take further diagnostic work up and another AI model which can pretty much work independently just in the hands of a healthcare worker.

Anasha Worker, who may not have a lot of skills from a clinical perspective, but she is the one who will communicate to the end users.

So instant triaging for one side and detailed reporting on the other side.

There are several things we Learned, but happy to share more.

Alright, I would also like to ask you, what is the challenge that one faces in terms of scaling up AI applications like Niram Dot AI with the government because clearly if you want to provide these applications at scale in different languages.

You also need the government as a stakeholder.

So what is the challenge that companies like yours face?

Yeah, you're absolutely right.

Especially in country like India, it's very, very important to partner with the government to reach to a lot number of women and scale up.

So in this case, I think one of the main challenge we face is that health is a state decision making, you know, entity, right?

So basically health decisions are made at the state level and very little can be controlled at a central level.

So we need to go to every state, convince the stakeholders, including, of course, starting from the health minister or ending at the health minister, but all the people in the chain.

And so it does take a lot of effort to convince them that AI can actually be much better than not having a doctor at all, right?

You know, having an AI enabled tool for, let's say, a triaging, instant triaging to identify high risk women or high risk patients in the remote areas and bring them to a more detailed work up in in the urban areas where the infrastructure is available is definitely feasible.

So we have to do a lot of feasibility studies, pilots almost for every state, and that's usually free of cost.

And it becomes a little bit, uh, harsh on a startup like us.

So to everyone you have to prove that it works.

And after that, we need to figure out what is the pathway to adoption.

So we're still sort of working on that level, but we've had a fair amount of adoption in Punjab, Maharashtra, a little bit in Karnataka and now in Harayana and so on.

Right.

I would also like to ask you, uh, what are some of the trends that you see across the world with new patents that are being filed, uh, in terms of AI for healthcare users, what are going to be some of the breakthroughs that we should watch out for.

Over the next two to three years.

Yeah, it's a great question because whenever we talk about AI and future of AI and and where it goes, we seem to be focusing more on the infrastructure level, like, you know, how can we make.

Of course, infrastructure is important.

We have to make fast decisions.

We have to have sort of deep Learned networks like LLMs and GPTS, etcetera.

Those are important, but those are not sufficient enough, right?

So I think the patents that we need to do more and so far we've filed and granted 39 patents, including 15 granted in the US.

So here we are focusing on what are those new biomarkers that we need to extract from this data that we already have so that we can personalize each of this diagnosis, screening, treatment, what say what, what.

What have you, depending on the application.

So how do we almost look at the DNA or radiomics extracted from AI from this data?

That's where I think we need to focus more patents on and that's where our initial patents have been.

And it has been, you know, able to trigger and fuel the breakthrough innovation that we have done.

But most of the focus is going towards infrastructure.

And yes, you need the machines, we need the platform, but the healthcare application is, is extremely complex because number of people who are healthy is much, much more the number of people who are unhealthy.

But unfortunately, you have to find those unhealthy women or men as early as possible and as accurately as possible.

So that is where the domain specific innovation is absolutely needed and new patents should be seen.

All right.

My final question, what is the kind of conversation you hope will take place on healthcare at the AI Impact Summit in New Delhi?

And what is the one thing we need to do to provide an enabling framework to companies like yours?

Um, I am really, really excited to be part of the um AI Impact Summit and I'm so happy that it's happening in India because India is going to define the future of AI, which is impactful.

Um, and so, so what I want to showcase there is definitely our solution.

We've got invitation to put our solution, a demo case in multiple stores and we will put in there and showcase to the rest of the world how we can really use AI to touch so many lives and save so many lives that we have already done so. And.

With this, we want to take our make in India solution to several similar low and middle income countries where you see a screening rate of 2%, like 98% of women are detecting their cancer by hand.

So I feel that this platform will do two things.

One, put India on the top of the AI impact country.

I would say leader.

Secondly, it will enable several of our startups like ours to reach out to similar countries and expand our reach beyond India and, and impact so many women around out there.

All right, thank you very much, Miss Manjunath for joining us with your experiences with Niram Dot AI and what lies ahead for AI in healthcare, we now shift our focus to NESA AI, a Mumbai based startup founded in 2023.

The firm provides AI acceleration cloud infrastructure for enterprises in India.

It offers platforms like velocities for GPU based training and inference, sizing Sovereign AI with open weight models.

It ensures data control, cost efficiency and compliance over global hyperscalers like AWS or Azure.

Joining me now to discuss the future of AI in India is NASA Dot AI's founder and CEO, Sharat Sanghi.

Sharat, let me start by asking you, how does the Union Budget help companies in scaling up AI infrastructure?

So, Parikshit, first of all, thank you for having me.

And I think the tax holiday that the government announced in the Union Budget for global cloud firms to set up infrastructure in India helps a new cloud provider like us because we can now be in a position to cater to their workloads in India, giving us the scale that we need. Uh, in addition.

To that, uh, India's, uh, the union government has also reiterated their commitment for, uh, budget for the AI mission.

So I think both these announcements actually help us because they allow us to invest in, uh, the AI compute network and storage, not only for the Indian audience through the air mission, but also for the global cloud service providers.

Also, how do you see the funding pipeline for AI companies in India?

So I think NASA may not be the right example.

I think since I'm a second time entrepreneur and I had a very successful first company, I've not had it difficult.

Uh, but, uh, and so we've, uh, been able to raise money from market investors, whether it was Nexus Matrix or Entity BC.

And we're about, uh, hopefully next week we'll announce a larger funding round.

But uh, I think, uh, in general, if for startups in the AI space, if they've been able to demonstrate, you know, client success and they've been able to demonstrate something that's unique, you know, most VC firms and most private equity firms are looking for, for companies.

Startups in this space.

So virtually every single major VC firm wants is looking at startups with some proven track record or with some kind of customer validation in the AI space. So it's actually most of the fun.