From Fintech to Farmlands: How Shubhangi Bansal Is Fixing the Trust Gap in India's Exports

India grows some of the world's most sought-after natural ingredients. Turmeric, spices, herbs, wellness extracts, the kind of raw material that global brands are actively hunting for. Demand is rising every year, and on paper, this should be one of the easiest wins in Indian trade.

It isn't. Not because the ingredients aren't good enough, but because the paperwork behind them barely exists.

Shubhangi Bansal spent a decade in fintech and trading infrastructure before she decided to go build something she could actually touch. Today she's the co-founder and acting CEO of Ikak International, a sourcing business that exports Indian turmeric, spices, and wellness ingredients to buyers across the Middle East, Europe, and the US.

Her days now look nothing like her old ones. She's on calls with manufacturers in Gujarat and Kerala one hour, chasing lab reports the next, and explaining to a buyer in Europe why last season's batch won't match this season's exactly. It's slow, unglamorous work. But she says it's the only way to build the one thing this entire industry runs on: trust.

We talked about why India's export documentation is such a mess, how she's using AI to run a very traditional business, and why she thinks agentic AI will never fully replace the human at the closing table.

Key Takeaways: What It Actually Takes to Export India's Ingredients to the World

The Documentation Problem:

• India has the farmland and the raw material, but most exporters don't maintain full traceability records the way buyers in other countries expect
• Every season produces a slightly different product, but many manufacturers try to make one lab report work across multiple batches to save cost
• Record-keeping is still happening on physical notebooks at the manufacturer level, which means the data eventually fades or gets lost entirely

The Buyer-Side Knowledge Gap:

• Some international buyers are highly technical and reject shipments over a 0.1 to 0.2 point specification gap
• Other buyers only know the price they want and a rough sense of color or size, with no real grasp of chemical or microbiological parameters
• Distributors and traders in the middle of the chain often understand the specs even less than the end users they're buying for

Where AI Actually Helps (and Where It Doesn't):

• Shubhangi uses AI daily for buyer research, validating company information, and running cold outreach at a scale she couldn't manage manually
• She believes agentic AI can fully own the monotonous, repetitive parts of a deal: sharing pricing, specs, and catalog information
• She's firm that the final close, where real trust and real risk get decided, will always need a human in the loop

Q: You spent years in fintech and trading infrastructure. What actually pushed you to leave that and start Ikak International?

Shubhangi Bansal: So there are like two, three levels into thinking why we switched and why I came into this journey. First of all, my father is also into business, he's a businessman, so that already was running into my bloodstream somewhere. But I found my own way, my path to groom myself, to get the confidence to do something.

After years of working in the corporate world, and actually working in fintech, made me realize that I keep working with the money which moves across borders but I've never been into something which I actually can see and do operationally. And this has huge global potential. Even our government keeps taking initiatives, keeps telling people that you need to export, you need to do things that are in global demand, and India has so much to give.

And third is the name itself. Ikak in itself means visionary. I had this vision that taking our Indian core ingredients to a global platform, though this already exists, what I feel I will bring into this journey is my 10 years of corporate experience, with which I know things can be automated. It's a traditional business, so I want to build this traditional business with today's modernization and continue that agenda with more clear intention, energy, and make it global.

🔥 ChaiNet's Hot Take: She didn't leave fintech because she was bored of it. She left because she wanted to hold something real. That's the difference between switching jobs and switching lives.

Q: When you actually got into the export trade, what surprised you the most about how it works?

Shubhangi Bansal: There were many surprises, to be honest. It's not that easy the way I thought, it's not that simple. The major gap I would say, and the challenge, is though global demand is huge, we have huge potential on our farmlands, in our industry, the gap where it lies is the quality. And even if we have the quality, there is something called, oh, the words don't align, because we don't have proper traceability of things.

In India, majorly people don't prefer to have full documentation, they don't prefer to give full clarity. To my surprise, any country outside you go, you'll find each and everything documented so well, so well linked, that you just give one unique identifier and you'll get the whole record.

I'll tell you one thing. In Dubai, if you go to any pharmacy, any medical hospital, you just have an ID, you give it to them, they will fetch all your details, all your previous medical records. In India, people will not find any detail anywhere. It's such an important thing, like health and medicare, but you don't have such records. Similarly in the export world, oh my god, it's completely missing, and you have to chase people like anything for documents.

🔥 ChaiNet's Hot Take: A country can have the best soil in the world, but if it can't prove what's in the bag, that soil doesn't matter to a buyer sitting in Rotterdam.

Q: So it's not just missing paperwork, it sounds like the paperwork itself gets stretched to cover things it shouldn't.

Shubhangi Bansal: Every season has a different produce. So when you have a different produce, obviously it's like a lab report, a lab test, because every time you cannot ensure things will be an exact match. You don't get a report every time, people just try, because obviously I know it's an expenditure, it incurs cost, but people try to minimize the cost. So promise something, deliver something. And it's a real challenge.

🔥 ChaiNet's Hot Take: Reusing one lab report across three seasons isn't cutting corners, it's writing a check the crop hasn't signed yet.

Q: You mentioned there's a challenge on the buyer side too. What does that actually look like?

Shubhangi Bansal: There is another bigger challenge, even at the buyer side. Some countries' buyers are very well versed, they know what they're buying, they are very technically clear, they have clear specifications. I will not name the countries, but some countries' buyers, they don't know what they're buying. Even if I ask them what quality they need, they'll just know maybe one specification, maybe a color. They will not be able to go into the specification tactics based on which prices vary a lot. You cannot give a very good quality at a very cheap price.

It becomes very tricky to manage that situation and chaos, because there is unawareness at the same time, but there is some awareness that this quality they want at this price. Oh my god, it's a huge chaos to navigate on both sides.

Some countries' buyers are so knowledgeable that even a specification gap of 0.1, 0.2 points, they'll reject, they'll simply say no. Though again, as I said, it's batch to batch, season to season, things change and you cannot exactly ensure everything.

🔥 ChaiNet's Hot Take: A supply chain doesn't break at the weakest link. It breaks wherever nobody agrees on what "good" actually means.

Q: You're based in Dubai but everything is sourced from India. Walk me through what a typical day actually looks like for you.

Shubhangi Bansal: I have a team which spans in India and I myself am based here in the Middle East. On a daily level, it happens, it's majorly, I attend majority through calling. And my team helps me get the pictures or visits, and I also make frequent visits to Indian farms and my manufacturing units who are my partners, just to get the hunch of how things are happening on a day-to-day basis. Just frequent visits which help me keep the trace.

What majorly helps me get their trust is giving them samples with the reports, because I emphasize on reports, that is what I can vouch for, which I cannot vouch for with just a sample. I say, you can test the sample, but here are my reports, and based on this you can validate your results, get the clarity, check the product, check the lab reports, test yourself.

It takes time. In this field especially, things do not happen in just one day, two days, or even fifteen days. It takes months to actually get a real conversion. It takes continuous effort and continuous trust building. Once the trust is built, and that too on the quality and what you're promising, the buyer doesn't want to go anywhere else, because they know it took time to build that trust.

🔥 ChaiNet's Hot Take: In export, the lab report doesn't close the deal. It just buys you the right to keep talking.

Q: You come from an engineering and product background. Has that background actually been useful in a trade business like this?

Shubhangi Bansal: I think there are two answers to this question. One, I would say, if I say I switched from being a developer to this profile, it's a big no, what you're doing there is no match, there is nothing common. On that front, it's totally true, there is nothing on the other side. I really don't bring any knowledge, I am literally beginning from scratch, I'm a zero here.

Flip side, there is definitely a lot of things that I bring forth. My 10 years of experience is my asset, my real asset. When I talk to people, like the owners directly, and I tell them about my past experience, then they look towards me with some credibility. I have earned my credibility and trust with that, because if any random stranger goes to you and asks for money, nobody trusts them, they don't have credibility in the market. So even though my export experience isn't enough, my credibility gives me that, buyers want to at least listen to me.

Second, because I am from a tech background, it instantly gives me the ability to figure out where I can do things faster, where I can automate. Probably these things might not click for somebody else, they might need a third person to understand those things. But to me these things click naturally. I have that clarity in my thoughts, because there is a logic in software, you build logic when executing anything in code. So whenever I do anything, my logic instincts work so high that I get to ask ten different questions before going to anyone else.

🔥 ChaiNet's Hot Take: Engineering doesn't teach you spices, it teaches you where the bottleneck is. Turns out that skill travels well outside a codebase.

Q: You mentioned automation. Where does AI actually sit inside your day-to-day operations right now?

Shubhangi Bansal: Yes, we run day-to-day operations with AI. It will be a lie if I say I don't use it, because with AI coming into prominence, it's a leverage that we take, it helps do our operations faster. With this I'm able to run scripts where I get buyer information from multiple resources, and I get to know what resources to use. I'm able to get the right information, segregate the fake information from the valid, and validate the information with their websites. I'm able to understand more about their products, their requirements.

On a day-to-day basis we do cold emailing, it's outreaching. For outreaching purposes I use AI. I cannot send 200, 300 people emails manually, it's not practically possible for me. So that's where it comes in handy. There are more systems I'll build eventually, but I'm still on a zero to one journey, and I'm still evaluating my current phase.

🔥 ChaiNet's Hot Take: She's not using AI to reinvent the export business. She's using it to send 300 emails she'd never have written by hand. Sometimes the win is just leverage on the boring part.

Q: Zooming out a bit, where do you see agentic AI actually changing how export buyers operate in the next two to three years?

Shubhangi Bansal: Agentic AI will help with the monotonous work, the every day monotonous work. So on a day-to-day basis, for any customer queries, for anything, even like chatbots we do interaction, with agentic AI literally a sales kind of thing can be started with that. Personally, at a later stage, a human person is needed because of the trust factor, you cannot trust a machine at the end of the day. I'm pretty sure in the next 20, 25 years, definitely AI will leverage things, it will make things go faster, but it cannot eradicate completely the human touch, feel, trust, a sense of humanity, emotions. For that, a human is needed.

In exports, majority of things is what you need to share, the pictures, the amount, the pricing, the specifications, these are all things where you don't need a person handling everything, it's a written thing, it will not change and it has to be there, someone in the system where you can just fetch and share further. So these things can be made faster and easier, but at the buyer side, at the later stage, a human would be needed for a longer lasting trust and relationship.

Where we need AI to help is bring a system into force where at any point of time they can see where their product lies, they can see the product journey, the commodity journey from the factory to their place. There's a lot more that can be done here, and it's a space that's very unoccupied right now. I would say it's a black hole and a lot can be done, but for that there's a lot to be done even at the Indian manufacturer side, because majority of details right now do not get recorded digitally. It happens on their notebooks, which is not a way to bookkeep, because eventually those papers get lost, the ink fades, you don't get to see the data long term.

🔥 ChaiNet's Hot Take: You can't build an AI-powered traceability layer on top of a notebook that's been sitting in a monsoon-prone warehouse for three years. The upgrade has to start on paper before it can start on a server.

Q: Do you think we'll eventually get to a place where a buyer's AI agent is talking directly to a seller's AI agent, cutting humans out of the early conversation entirely?

Shubhangi Bansal: I'm not sure if it's true or not, but we know the platform Alibaba, I feel there are chatbots on their sales team, and anytime you message, they respond immediately. So it's more like chatbots, initially what I feel is a chatbot has been integrated which sends automated responses. For initial discussion, sharing files, sharing information, sharing catalogs, asking normal queries about you, what you need, how much quantity, these are questions that are written all over there, for these things definitely, both sides, manufacturer and buyer, agentic AI can work.

But as you proceed higher, and where you have all these basic details with you, a human interception is needed, because a human eye is needed to evaluate what works for them. Maybe you give it to AI and ask which is better, but AI might not be right every time. You have to have human involvement to see what works and what doesn't, because at the end you're making a real life product which will be consumed by a real living being, either an animal or a human, and that needs to be validated by you. At the higher level, human interception would always be guaranteed and required, because I don't think complete reliance on AI can happen at any day.

🔥 ChaiNet's Hot Take: Agents can handle the small talk. They can't handle the moment where a wrong call means someone eats something they shouldn't have.

Q: You gave a great example earlier about a number rounding off and causing a health risk. Can you walk through that?

Shubhangi Bansal: Let's say I just imagine, by any chance, if it malfunctions, at the end of the day it's a machine, it can malfunction. Let's say something had to be five, it should not be more than five, and it came out to be 5.2. It can round off and say it's fine, it's five. But that 0.2 can cause a bad impact, anything you're going to consume, oh my god, it's a mess. It will cause a health problem. Machine cannot be trusted, simple. At the end of the day, it's been built by human for human help and leverage, not to give it complete control.

🔥 ChaiNet's Hot Take: A rounding error in a spreadsheet is annoying. A rounding error in a pesticide residue report is a recall. Precision isn't optional when the output gets eaten.

Q: Is there a product that's in high demand from your buyers that you're just genuinely struggling to source right now?

Shubhangi Bansal: Yes, it happens with me. Organic products are very highly in demand, very high in demand. And in India it's hard. The major reason is our soil, it's not about the way we cultivate, people are practicing better ways to cultivate, but the soil itself has grasped those pesticides, it has got changed, it's mutated, so there are traces in the soil itself. And that's why even if manufacturers are growing and practicing good cultivation, taking care of all the right steps, somewhere that trace causes rejections. We're not able to say the products are organic, though many manufacturers are still working hard with farmers on contract farming.

Some commodities I'm just not able to find. I really feel the heat and I have to really fight, but at times I fail, because even a slight trace will fail the product at the global market, and we can't take the risk, because these products, at literally the port, they fail in a test and get returned. So you can't take the risk.

🔥 ChaiNet's Hot Take: Demand for organic isn't the bottleneck. Decades of pesticide residue sitting in the soil is. That's not a marketing problem, it's a generational one.

Q: For someone outside the industry, how do you actually determine whether a product qualifies as organic?

Shubhangi Bansal: Different countries have different parameters, but the major one is pesticide residue, it should be pesticide residue free. And then different products have different requirements, for example there should not be ethylene oxide, it should be eto-free for European countries. There's a minimum residue limit for pesticides that has to be agreed upon or tested for, and even if it shows a slight variance, they get rejected. It's a real thing, and for that you literally have to find soils that have not been mutated, where you can still cultivate, but that area is smaller comparatively. We have a huge farming ground, but those clean areas have reduced, so that causes issues, we don't have a lot of product for supply. Demand is huge but supply is less.

I've actually said no to buyers before. I tried my best looking around everywhere, even with local vendors, thinking maybe I'm not able to find the source directly from the farmers, so let me try the local vendors because they're on the ground, maybe they know more, there are a lot of mandis and I might not be able to track each one. But eventually I figured out, no, I'm not able to get it, and I had to say no to the person after really putting my head into it for three, four days. Then I thought, there's a lot of scope in this, demand is huge, and if we're able to provide these materials we can really leverage it, because some things are only grown well in India, but we're not able to take that advantage.

🔥 ChaiNet's Hot Take: Telling a paying buyer "no" is one of the hardest things a founder can do. It's also usually the moment they realize exactly where the real opportunity is hiding.

Q: Your portfolio has six or seven categories, but right now you're mostly focused on agri. Why that discipline?

Shubhangi Bansal: We have other portfolios as well, the only thing is, because it's not been long, I'm still on zero to one, so I'm right now focusing on agri. If somebody asks I provide, but I am currently focusing on this, because you can't take everything and just be an expert on all of the things. You have to own one thing and be an expert in a way, at least so that if something's being asked, you can answer or ask ten things about it. I believe in doing that first, that's my agenda.

🔥 ChaiNet's Hot Take: Six categories on a pitch deck looks impressive. One category done with real credibility is what actually gets a buyer to sign.

Q: Last question. Looking three to five years out, with younger operators and better tools coming into this industry, do you think India can beat China in exports?

Shubhangi Bansal: I would say we should think of it in a way rather than beating any country, we should learn the way they're operating, and definitely taking those cues from them, we should be able to uplift ourselves, because we have so much that we don't need to beat anyone, we'll just grow a hundred times, I'm telling you, and automatically we'll be number one. Everybody knows we are the top most country in population, the major asset we have is our human population. We need to leverage that. If we can leverage our human population and make them aware of what we have today, AI is something we can leverage, we have people resources, we have AI, we have our agricultural land. Using all of these we can make our country grow a hundred times. The only thing is, unite all of these factors, make the best use of these available resources, and we can aim to be number one.

🔥 ChaiNet's Hot Take: She's not chasing a rivalry, she's chasing a ceiling India hasn't hit yet. That's a much bigger goal than beating a competitor's number.

Final Thoughts: The Real Bottleneck Isn't Demand, It's Documentation

Shubhangi's closing perspective: "We have so much that we don't need to beat anyone. We'll just grow a hundred times."

The bottom line: Every conversation about India's export potential eventually lands in the same place: the demand is real, the raw material is real, and the gap sitting in the middle is trust. Not trust in the abstract sense, but trust built out of lab reports, traceability records, and consistent documentation that most manufacturers still don't produce. Shubhangi's biggest surprise moving from fintech into agri exports wasn't the industry itself, it was how much of the friction had nothing to do with quality and everything to do with paperwork nobody had bothered to standardize.

What's interesting is where she draws the line on AI. She's not romantic about the technology and she's not scared of it either. She uses it every day to run outreach and validate buyer information at a scale a small team could never match manually. But she's blunt that the final handshake, the moment a buyer decides to trust a shipment they haven't seen yet, still needs a human who can be held accountable if something goes wrong. Her rounding error example, a 5.2 that gets reported as a clean 5, is a simple way of making a serious point: in a business where the product ends up being consumed by a real person, a small machine error isn't a rounding issue, it's a health issue.

For anyone thinking about building in India's export or agri-tech space, the opportunity she's describing isn't really about finding more buyers. It's about building the infrastructure, digital records, traceability, consistent testing, that lets Indian exporters prove what they're already capable of producing. The demand is already sitting there waiting.

Q: How can people connect with you and learn more about Ikak International?

Shubhangi Bansal: Thank you so much for having me here, it was a wonderful conversation. People interested in Indian ingredient exports, whether they're buyers looking for sourcing partners or people curious about what's happening on the ground in India's export industry, can reach out to me directly and I'm happy to share what I've learned along the way.

Final words: India doesn't have a supply problem or a demand problem. It has a trust problem, and trust gets built one lab report, one documented batch, and one honest "no" at a time. Shubhangi Bansal's bet is that the founders who fix that unglamorous middle layer, not the ones chasing the flashiest AI headline, are the ones who'll actually take Indian exports global.

---