Pod Insights The Brand Called You Empowering NGOs Through Responsible AI: Jyothika Raju, Co-Founder & Program Director, ImpactAI Foundry About this episode Join us for an insightful episode of The Brand Called You as host Ashutosh Garg speaks with Jyothika Raju, Co-Founder and Program Director of ImpactAI Foundry.
Discover how Jyothika's journey—from studying engineering in Bengaluru to leading initiatives at the intersection of artificial intelligence, public policy, and social impact—has shaped her vision for responsible AI.
In this conversation, she explores:
What responsible and accountable AI truly means
The challenges NGOs and nonprofits face when adopting AI
Why capacity building is more valuable than technology dependency
Practical AI solutions already transforming the social sector
The importance of collaboration between governments, universities, startups, and nonprofits
Actionable advice for young professionals pursuing careers in AI for social good
Whether you're passionate about artificial intelligence, nonprofit innovation, public policy, or ethical technology, this episode offers valuable insights into building AI that creates meaningful and lasting impact.
Transcript 0:08 Speaker 00 Welcome to another episode of the brand called You, a podcast and podcast show that brings you leadership lessons, knowledge, experience and wisdom from thousands of successful individuals from around the world. I'm your host Ashutosh Garg and today I'm delighted to welcome an accomplished leader with a deep understanding of AI from Bangalore, India, Jyothika Raju. Jyothika, welcome to the show.
0:31 Speaker 01 Thank you so much. It's such an honor to be here.
0:33 Speaker 00 Thank you. Jyothika is the co-founder and program director of Impact AI Foundry. So Jyothika, let's begin with your early years. What first sparked your interest in technology, public policy and social impact?
0:50 Speaker 01 So I think there was like one moment as such, I think when I was growing up, obviously when you're in Bangalore, I think engineering is the thing that you're thinking about all the time. And I really didn't have a specific goal or I didn't know what I want to do in life yet. And I was just very open to exploring and seeing where I could fit. And I think computer science was something that interested me, but I didn't know if that was what I wanted to do for the rest of my life. And obviously I'm like talking to family and being like, how do I figure this out? And my dad was like, you know, engineering is the most like versatile thing. You can do anything you want after that. So that's how I went into computer science. I did my undergrad in Bangalore. But while I was studying
1:34 Speaker 01 and doing like working on projects and things like that, I realized that a lot of my projects tended to do with social impact. And I also started volunteering in Bangalore with a couple of NGOs, which is where I think my interest in social impact comes to. But what really clicked for me over time is that I I wanna find out how to meld both of my interests like tech and um impact and that's where I think policy came in where I didn't want to just work in the nonprofit space or just work as a software engineer, I wanted to see where most structural change could come in and like try to really um effectuate how uh like change conditions for everyone who are working in the system, right?
2:18 Speaker 01 And yeah, I think that's where like po- policy seemed like a glamorous choice.
2:24 Speaker 00 Yeah.
2:24 Speaker 01 Um I'm still in doing my masters right now, so I'm not sure what the reality would look like. But yeah, I think I'm still finding out a space to bring all of these three things together.
2:34 Speaker 00 Fascinating. It's amazing how you found literally a cross section uh of all these three between technology, policy and impact.
2:43 Speaker 01 Yeah.
2:44 Speaker 00 Very exciting. And then you have managed to bring in the overall excitement of A I into this uh triangle as well. Tell me a little bit more about your interest in A I and how do you personally define responsible and accountable A I?
3:02 Speaker 01 Yeah, I think um of I mean, when I was in college, which was like like two, three years ago, I think that's when A I became a really big part of the conversations that when Chat G P T came out, then Claude came out and if you were an engineer, I think um interacting with A I became inevitable.
3:19 Speaker 00 Yeah.
3:19 Speaker 01 And that's where you sort of uh interact with it, figure out how these tools are working and how you can po- po- possibly sort of further your interest in that. But I think I also started noticing what the problems could be with A I itself. Like I think when we first saw all of these models come out, the thing that people were grappling with most was bias, I think. These systems were biased, even in things like facial recognition and hiring algorithms. And I think even now, recently, we see um a lot of bias in these L L M's and that's where I really started looking into responsible A I and like really understanding what that means.
3:59 Speaker 01 And I think for me, responsible AI really um is about who is sort of building these tools out and who your end users are and how you're sort of using that data to not further like the discrimination that we see already. And I think because of the data that these systems are sort of trained on bias, um inheriting bias is inevitable.
4:22 Speaker 00 Hmm.
4:22 Speaker 01 But I see that um to really further responsible AI, I think governments need to come into the conversation. I think a lot of countries, I think even South Korea and Japan have a lot of AI legislation that's being passed.
4:35 Speaker 00 Mm hmm.
4:35 Speaker 01 Right now and um even really understanding that there are very few companies that are sort of controlling what the conversation around AI is looking like.
4:45 Speaker 00 Hmm.
4:46 Speaker 01 And really um I I like looking at how you can shape this ecosystem, not just as like a voluntary ethics commitment on their websites, but more structurally like the EU AI Act is like an example of how you can really mandate like fairness audits and things like that.
5:04 Speaker 00 Hmm.
5:04 Speaker 01 But when we were building like Impact AI Foundry, which is the program that we're running, I think when we look at, we're really thinking about being very responsible with AI, especially because of the communities that these um NGOs are going to be working with.
5:17 Speaker 00 Hmm.
5:17 Speaker 01 And I think there are three things that I personally sort of look at for responsible AI. The first is like honest representation, like really understanding who these communities are and who your stakeholders are in the design process itself and not just as an afterthought of these tools itself. And the second is what's the honest capability, like what can this AI tool actually do do for you? Like AI shouldn't just be used because it's a buzzword right now, you shouldn't adopt it and just like put it into wherever it can fit. But really understanding what can this tool do for you.
5:51 Speaker 00 Hmm.
5:51 Speaker 01 And I think accountability is also a big part of the conversation where you need to have a human in the loop, especially for the context of where these organizations work in.
5:59 Speaker 00 Hmm.
6:00 Speaker 01 Where you're able to intervene and I think that's what sort of separates responsible AI from the whole like overarching conversation about AI and yeah, that's.
6:09 Speaker 00 Great response, great response. Thank you. Do you think, tell me, what did your work in AI governance research teach you about the opportunities and risks of AI?
6:20 Speaker 01 Yeah, I think there are a lot of opportunities, especially for in the space that we're working with NGOs. When you have a lot of resource constraints, you have a lot of, you have tight budgets, you can't really like hire a lot of technical talent. I think there's definitely an opportunity for people to upscale. I think AI is making a lot of things more accessible for people who are not engineers. Like you can create your own website now, you can create a small app by yourself, you can do a lot of these things. But I think the risks are also real, like I was talking about things like bias, things like accessibility. And also, I think in the global south, because a lot of these systems are designed with a very Eurocentric viewpoint.
7:01 Speaker 01 I think it becomes difficult to adopt them for people who don't, who speak the language that these models were not trained on. And I think we need to be very careful about who AI is going to be affecting the most and who's like the people that are vulnerable. Like I see it as like, I think a lot of government governance frameworks are trying to look at it as a very neutral infrastructure. But even if you consider AI being like a commodity like electricity, I think even electricity access is very unequal, right? And I think it's important to be very deliberate when we understand what are the opportunities and what the risks that come with it itself.
7:41 Speaker 00 That's right, but since you spoke of electricity, I mean, you know, energy is directly and yet indirectly related to AI. Because of the immense need for all these big machines for electricity. But then what lessons from the sustainability sector have influenced your work in technology policy?
8:04 Speaker 01 Um, yeah, I think my first sort of twist with the sustainability space came with a think tank that I was working with. And there I really was working on like a renewable energy citing framework, like to find areas where we can really build these big solar farms and renewable energy plants in general, right? And I think the question there was how do you sort of deploy something that is genuinely like use? Full for like the public itself, but really without destroying or affecting the communities around you that you're actually trying to protect, right?
8:38 Speaker 00 Hmm.
8:39 Speaker 01 And when you're mapping this, you look at things like environment sens- sensitivity, you look at social vulnerability and you look like how these li- livelihoods of people around you are sort of getting destroyed, right? And I think that's the word for word um challenge in A I itself, like it's been very slow to adopt like precautionary thinking they're more reactive than proactive.
9:01 Speaker 00 Correct.
9:01 Speaker 01 So I think even if you look at A I data centers, right? I think the energy demands of that is quite high and they are the largest consumers right now of electricity, water. And I think another um concerning fact is that a lot of these data centers are moving into the global south. Which means that there's already lower protections for people here and it becomes even more important to sort of get more regulation in place to make sure that these vulnerable communities aren't getting affected. And also because this is when you think about renewable energy and if you compare it with like what's happening with A I, R E itself, like when you
9:40 Speaker 01 take a new swath of land, it's not just a techn- technology swap, it's just you're reshaping how the land is getting used, you're replacing supply chains and you're also like reinventing what the labor market is going to look like for that community itself, right? And I think um that's what A I is going to be like right now and like the risks are really real and I think we all need to be grappling with that right now.
10:04 Speaker 00 Well said. Do you think you're the founder of Impact A I Foundry? Tell me about Impact A I Foundry and what gap did you notice in the nonprofit ecosystem that inspired you to create this?
10:17 Speaker 01 Um. So I think Impact A I Foundry, um just to tell you the bare bones of what we're doing is uh we run this program for two months where we have four in-person sessions where we bring a couple of NGOs into a room together.
10:32 Speaker 00 Hmm.
10:32 Speaker 01 And we help them to understand, uh think about their workflows, think about where A I can help you, um what can you automate? What's better left not touching and just basically helping them think through that and building capacity. And the way that we do this is that we pair them with like an engineer from Bangalore, someone who has like a deep knowledge of how to use AI, how to use technology and what kind of problems can be solved using these tools. And the reason that we do that is um we don't ask these technologists to build for the NGO. We want the NGOs to adopt these tools and to understand how to build them themselves. So it's more about um trying to help them walk away with a working tool.
11:17 Speaker 01 But also making sure that they are going to be able to build for themselves rather than relying on like an engineer or someone that they have to hire or spend more money and funding on creating these new tools itself, right? So that's the aim that we're we really have when we start out this program. And when we look at um what gaps are in the nonprofit space right now, I think um the the most stark one is obviously things like funding, like budgets and time and capacity building, all of these are barriers, right? Um it's not that NGOs don't want to try it, but there's always a sense that you know you have to spend more time on it, you have to spend money on it and it becomes difficult to sort of
11:59 Speaker 01 um really uh go ahead and see how this can help you. It's difficult for them to be creative. Uh when they are already constrained. And through like when I was volunteering to this, that's what I noticed that there were a lot of issues that you can solve, but you don't have the people to solve them or you don't have the know-how. I think even the talent that gets attracted um maybe not the best people to help you build these tools. So I think for us, what we want them to walk away with is not just something generic and not like an AI training where you're just learning what this is and um things like that, but more that like you're able to build a to- tool that can become something that you yourself built, you maintain it and you adapt it.
12:43 Speaker 01 So it's more of a learning and a teaching process than actually like it would have been easier for us to go in and build this tool for them. But I think that's not helping them or it's not sustainable in the long run.
12:55 Speaker 00 Correct.
12:55 Speaker 01 So yeah, that's what we're trying to do.
12:57 Speaker 00 Well said. But you know, given the um availability of AI in virtually every. Company, every human being is talking about A I. What are some of the challenges nonprofits face? When it comes to adoption of A I, other than just, you know, going to a chat G P T and saying, uh, help me write an email or or get me a a new I T ready.
13:24 Speaker 01 Yeah, I think um the challenges are a couple, right? Firstly, again, funding, even to get your staff to get trained on how, um, everyone can go write a generic prompt, I think, but to really understand what these tools can do for you and how much of value you can extract for them, you still need to have a base level of training on how to exactly do this.
13:47 Speaker 00 Hmm.
13:47 Speaker 01 And I think that's where funding becomes a problem because all our resources are dedicated to working on the ground, especially if you're very grassroots organization. I don't, I think you're always going to grapple with how do we, where do we put this money to, right? And I think the second problem would be um attracting like the right talent to, right? I think it's not because the smart people don't want to come into the nonprofit sector. It's just that there's no real pipeline, like there is no way for an engineer in Bangalore to be able to interact with an N G O and actually go build something for them. But um and even if they build it for them, it needs to be like a continuous process. You can't build something and leave them to sort of handle it because then that just breaks down.
14:29 Speaker 01 And I think um really uh building people who know the problem need to be building. And that's what the gap is, right? Like you can be in a like a a for-profit organization saying we'll go in and build this for you, but you don't know what the actual problems are and building trust, like being responsible about how you build um these systems. But also really um like nonprofits have a lot of data from all of these vulnerable stakeholders that they work with. So how do you protect that data? How do you get them to trust you to trust these tools itself? I think there are still a lot of people.
15:04 Speaker 00 Correct.
15:05 Speaker 01 Who struggle with that.
15:06 Speaker 00 Well, sir, can you give me an example of, you know, one or two of the kind of A I tools that you have uh helped build?
15:13 Speaker 01 So this program is actually, we derived it from a program in New York and I actually volunteered as a technologist there too. And I think there were really interesting things that the nonprofits built out there. I think one was like an adaptive teaching platform for students where it would ask you a question if the student put in the right answer, it then gives you like good and takes the next question. If you don't understand it, you go back, you you understand what level that student is at and then give them an explanation that's appropriate to where they are and really helping them understand what's happening, right? And I think automating administrative tasks has been a big thing, like you know, create like getting grants into a database, sending out
15:58 Speaker 01 you know, grant writing and reporting. And I think a really important and useful tool that nonprofits are able to create is like data dashboards. Like a lot of them have all of this data, they have field data. But just to be able to visualize it in a more easy manner, even for a non-technical team member to be able to really get an interactive dashboard out there, I think it's become a lot more easier to do that. And things like chatbots, like multilingual chatbots and these kind of different use cases that are very context specific. Like we want them to build for what's the context of their organization and what they really need.
16:37 Speaker 01 Rather than us saying, you know, build this out as an example, we ask them to come with their problems and then figure out what they need.
16:44 Speaker 00 Hmm, well said. How do you ensure that the AI solutions that are being built for nonprofits or that they are building themselves, they remain ethical, transparent and genuinely useful?
16:58 Speaker 01 Yeah, I think that's built into our, the way we teach itself is because firstly, I think the first safeguard to really understand if this is a useful tool for you is to start with the workflow, not with the technology or with the tool. We're not going to be like go and start using like Claude code or something like that. We spend a significant amount of time really helping them map their existing processes and see what is it. That you can automate with A I, where is it genuinely going to add value? Like, we just want them to keep asking questions of, does this add value to us? Where can it add value? And does it, is it going to help us reduce time? Is it going to help us save money? You really need to be asking these questions to know whether this is a useful tool for you or not.
17:40 Speaker 01 And even if it's a slower start, I think in the long run, it really helps you um and prevents a lot of harm. I think the second question we ask is for them to think of who they're working with, what kind of data they're using. What are you building and who is this going to affect? Like, if something happens and it fails, what's the? Who is this going to affect and who is the person that's going to be harmed? And I think when these conversations keep happening, then they're able to think of all of this when. You're designing it itself and not after something happens. And I think the third is like um like independence, right? The thing that we want to give them is independence for them to be able to own it. And for them to be responsible and accountable to maintain it.
18:22 Speaker 01 So, yeah, I think that's sort of built into the structure of how we're going to be doing the sessions.
18:27 Speaker 00 Very interesting.
18:28 Speaker 01 Oh, yeah.
18:29 Speaker 00 What have been your biggest lessons as you have built Impact AI Foundry?
18:36 Speaker 01 Uh, I think the first thing that um I learnt was that um. As someone with a technical background, you always want to move fast. The minute you see a problem, you're like, okay, we can like create like an app to to solve this. We can create like. Um, this technology to solve this, we can use this to solve it, but. Really like stepping back and asking that, does this actually work? Is this actually going to help you? I think really getting slow and and sitting with the process is uh it was like uncomfortable at first, but it becomes more easier when you realize that. For something to really help, you need to go slow and you can't just like solve every problem with a tech. Um, you know, tech perspective.
19:14 Speaker 00 Hmm.
19:14 Speaker 01 And I think the second thing was um like the difference between us being helpful versus us, like the nonprofits getting dependent on us. I think it it's very easy to become like a tech consultant and be like, okay, we'll build this out for you.
19:28 Speaker 00 Correct.
19:28 Speaker 01 But teaching is a much slower, it's a mess messier process than actually building it ourselves and. I think really the thing that we're focusing on is how to like have that patience to really get that foundation in for these nonprofits. And I think maybe the third thing, which is a little smaller, but one more thing I'm learning is because it's just a very small team running this. It's about, um, you know, switching between the operational part of it to the visionary part of it. Like both are these, both of these are your jobs. So yeah, I think really like balancing writing a catering plan on one day to really like envisioning what the future of the organization is going to be. Both are happening at the same time. So, yeah.
20:10 Speaker 00 Great response. Thank you. A time for two more questions for you. Um, as you look ahead, what role do you think governments, universities, startups and nonprofits must play together to shape the future of AI?
20:25 Speaker 01 Yeah, I think it's a very interesting question because I think every, all of these stakeholders come with their own, um, like they have something real to contribute. But there's also like a genuine blind spot that can come in the way, right? I think governments have a lot of reach and legitimacy, but they move a little more slowly because they have to be careful about everything that they're doing. And universities, I think, are even though they produce a lot of knowledge, they're dispensing knowledge, they're very disconnected from the real implementation part of it. And I think startups optimize obviously again for like growth and for profits and they're not always aligned with like social impact and those kind of goals.
21:04 Speaker 01 And even though nonprofits are the ones closest to the communities that they're trying to help, the communities that are affected, I think they have resource constraints. We stop them from doing a lot of things that these other organizations can do. So I think there's a genuinely a need for collaboration and like really understanding how you can structurally tackle this because it's still a new frontier, right? AI is still new, you can really, um, you can hard use it for, um, like doing a lot of good. And I think it's good, it's a good first step for governments to really understand how they can step in for universities to, you know, create partnerships and really startups and NGOs to work together to sort of.
21:49 Speaker 01 And that's what we're trying to do here to like bring the for-profit sector and the nonprofit sector together to be able to do something that's going to help a lot of.
21:59 Speaker 01 Yeah, I think it's the moment that matters for us now.
22:02 Speaker 00 I agree. And my last question to you, Dottika, for young professionals like you who want to work at the intersection of AI, policy and social impact, what advice would you offer them?
22:15 Speaker 01 Um, I think I'm also still in my early in my journey. So rather than giving advice, I would like actually be, uh, I would want to share what's working for me right now.
22:25 Speaker 00 Yeah.
22:27 Speaker 01 The first thing I would say is that if you're someone that's like interested in social impact, really stay close to the people that are affected by the problems that you care about, right? Not just, um, like, just so that you understand what the real issues are and not from your perspective, but from the people who are actually going through these, um, problems. I think it's, they have the best lens of what they need and, um, it's better for you to be an observer and then work on it rather than thinking that you have all the best ideas. Because I think even for us, the best ideas haven't come from feeling like we know everything, but really sitting with like NGOs, they're sitting with people
23:11 Speaker 01 and really understanding what are the problems that you're dealing with and how can we change this for you? And maybe another thing I try to say, and even I'm trying to figure out for myself, is don't try to, like, there's a lot of pressure to specialize really early and really find that spot where you'll like create your own journey. I think this intersection is new, like AI plus policy plus social impact is still a new territory and I think the people who are doing a lot of work here don't fit into any existing categories itself. Like it is a bit of an ambiguous space. You don't know where it's going to go tomorrow. So I think it's, it's a good place to be in because it's going to help you really think about what is it that you want and
23:55 Speaker 01 what's going to make you the most happy? Is it going to be working with a grassroots organizations and building our tech for them? Or is it going to be like working with the UN and creating like governance frameworks for them? So yeah, I think keep asking yourself what you want and it's okay if you don't have the answers. You will figure it out.
24:13 Speaker 00 Absolutely. Or you can ask AI to give you the answer.
24:16 Speaker 01 Definitely, yeah.
24:17 Speaker 00 But on that note, Jyothika, I just want to say what a fascinating conversation this has been. I mean, quite an all encompassing discussion on AI, yet highly focused on the cusp of technology impact in social sector. Thank you for the amazing work that you're doing, supporting so many social sector organizations through Impact AI Foundry. All the very best to you and good luck.
24:41 Speaker 01 Thank you so much. It's a great conversation. Thank you so much.
24:46 Speaker 00 Thank you for listening to the brand called You videocast and podcast, a platform that brings you knowledge, experience and wisdom of hundreds of successful individuals from around the world. Do visit our website www.tbcy.in to watch and listen to the stories of many more individuals. You can also follow us on YouTube, Facebook, Instagram and Twitter. Just search for the brand called You.
Previous Thorsten: Venture Capital Leader & Innovator | From Deutsche Telekom to the NATO Innovation Fund & Thinkstorm Ventures Next Empowering NGOs Through Responsible AI: Jyothika Raju, Co-Founder & Program Director, ImpactAI Foundry Page updated November 14, 2026 · transcript and summary are machine-generated