Trustless Verification for Financial Institutions - Kofi - zkTLS Day @ Devconnect 2025

Transcript

How's everybody doing? Good. Nice. Um, all right. Thank you guys for coming. My name is Kofi. I am the co-founder of Credentials and at Credentials, we build verification infrastructure for financial institutions with ZKTLS. Um bit about my background uh is in cyber security mostly uh NIST sock 2 data privacy GDPR. So yeah I'm just going to take you guys through the problem that we're solving um how we're creatively doing it with zero knowledge and the type of access it's able to unlock globally.

So there's a huge problem within the global digital economy. Um millions of people are making money online. uh yet traditional and non-traditional rails uh are built properly to like accept this data. So if you're somebody that's making money from like the gig economy, so like Uber, Upwork, or you freelance or you're on YouTube and you want to get access to like a loan, a house, um all this remains invisible from like traditional financial systems today. uh mostly kind of like how Ricardo said, they don't necessarily take PDFs or like screenshots as verifiable data um for something to assess risk on whoever is like coming to get financial services.

Um a highle overview about this is like data itself on the internet is fragmented by throughout thousands of platforms and there isn't like one platform that has like a unified source of truth for um banks to trust for neo banks to trust um for people to do on and off ramping. Um quick story I can sure like share about like how like financial fraud is within like the traditional banking sector. Um my dad has like a history in oil and gas. Um created his wealth that way out in uh West Africa, Ghana. Um one day the minister of finance came to my dad for a loan of $500,000 um guaranteed with like one of his estate properties. Um because open

banking doesn't exist in Ghana. The minister of finance also went to a bank um a Stanley bank requested the same 500k. um the bank wasn't able to see into like his finances to understand that he had already taken out credit from another person um which resulted in him burning both my dad and the bank right so um fraud is very high in places where data is invisible or like being able to see um people's actions and such um but in 2025 is actually pretty surprising how we don't have like efficient systems um like twothirds 90% of the world has a smartphone um about over a billion people remain unbanked, but we all like use finances online, whether it's like crypto, whether you're

doing ride sharing, whatever the case may be. Um, twothirds of us around the world make or receive digital payments. Um, and yeah, like I said, this global digital economy that we're building um doesn't involve open banking for some reason. Like I can't take my credit score, my financial history from the US and go to Lisbon to go take out a mortgage because they don't necessarily have the rails to properly verify this information. Um and existing solutions that exist today uh are very limited. Um APIs, public APIs don't cover like revenue data for people. Um screen scraping is very risky um to institutions because you have to take the users like username and password stored in your servers. Um there's many stories like a couple

months ago Alliance Life which is like a life insurance company um got hacked, cyber attacked and bunch of user credentials were exposed. Bunch of data was leaked. they have to pay like millions of dollars in settlements. Again, like screenshots and PDFs are not they're not verifiable information for you to make decisions on. So, there's no like global standard for um folks to like lend um to move economies out of cash and carry to like lending and credit around emerging markets. Uh which is why we decided to build credentials which is a verifiable um authenticity layer for the internet.

Um using ZKTLS we can instantly verify incomes transactions uh without exposing people's personal information. Um it's universal coverage as well. So it works on any HTTPS dashboard wherever money exists online um to provide lending opportunities KYC digital nomad visas and insurance um and also automate crossber payments. Uh so yeah so we support many different platforms uh currently mostly like gig um economy platforms eventually mobile money um social platforms crypto exchanges and also like financial banks locally in different regions uh I think I'm going to demo now uh time

is okay so we usually have a dashboard uh or APIs for like banks to use and what happens is you can use these to send out verification links so imagine like you get a link on your WhatsApp or your email for you to complete a verification of where you make money. Um, you're sent to the credentials application and that's when you can either select one platform, multiple platforms. This is going to give an example of Binance and it basically if anybody here has used Plaid before, it's kind of like the same UX. You get navigated to then log to your um source. Um, so right now it's it's getting ready. So yeah, you log in, put in your email and password. Um, we'll let this run.

Yeah, but basically once you've logged in, we're then able to like collect um data. We're built on top of reclaim. So we use like their zero knowledge engine to verify the data that's on the platform. This works with anything. Could work with a bank account, could work with your Uber account, work with your crypto platform, um to create a verifiable proof of like your income holdings, um so on and so forth. Yeah. Yeah. And then once this is now created, you then are able to like submit this. And this all happens locally on the user's device too. We don't see the data or store it at all. Um neither does like the company or like the lender. This is a bit small here, but like this is how the dashboard would look. You can then go and like view the uh verification

itself to see the person's like monthly income and like withdrawals. and using this as like a verified source of truth uh instead of like a screenshot you can now then deploy micro credit or loans uh auto finance mortgages depending on your risk appetite as like a lender um so that's one of our products uh that we have um the second one that we use so we usually uh support like neo banks banks and we also support on and off ramps um for like automated on and off ramping with our SDK uh we basically built using AI a headless browser and zero knowledge we're able to verify bank transfers. So, somebody comes to you and they want to onramp $10,000. Um, instead of like having to call up your banking partner

like say, "Hey, has the money restart our merchant accounts, you can then have this person store their um username and password in a config file, which you don't have access to. We don't have access to." And AI just basically goes in, logs in, verifies the bank transfer. Um like as you can see here, this is very blurry, but like it will have like it's verified the amount in the result and then it allows you to permissionlessly release stable coins from escrow um in seconds. So like yeah, as you can see here, it's checked out here of the transaction. Um uh so yeah, so what this unlocks is it allows for um like verifiable income for digital workers. Um, if you guys know a company named Safety Wing, they do digital nomad insurance um, for people

that go across the world and maybe like stalled out for work. One of the reasons why they wanted to kind of use our product, it was for the verifiability of people's earnings, right? They they're able to dish out insurance um, for guys that are stuck in between jobs. Like I said, faster on and off ramping. You can also use this stuff for like soft KYC and identity. Um, and to basically open up like borderless uh travel like within the digital nomad or like the digital economy right now like I myself can go around the world and take my credit history with me. I can take my finances with me to get access to financial services from across. Um, so that's that's kind of like the idea that we're executing on uh

within the real world. Um so far we have a couple of pilots running with a few like on and off ramps to using our SDK for like this automated P2P um verifications and then we also have like neo banks. Um for example, Cassie Money is a neo bank for the gig economy. Um they use us which we use ZKTLS to verify Uber income to allow them to dish out micro credit to drivers. Um I think I know a few guys know like Lemon from here is like one of the biggest like crypto latam platforms uh within this uh ecosystem. They have 5 million users within like the digital economy, creators, um gig workers, um local people that work here, but they have no

access into like their income levels. All they can see is people are depositing money, they're spending with their lemon cards. Uh but with credentials, they can now like increase people's transaction limits. They can also start lending to people here in Latin America. Um and which is very big too because over the last 25 years like depreciation on a dollar hit like 90%. Um this has become more of like a cash and carry economy. So now you can even create like verified credentials on Lemon which is like a real reputable company. Then go to another bank and say like hey I've verified my income with Lemon. I want to get access to said loan or like said house and things of that sort. Um yeah. So those are pretty much the platforms we support. Um and [sighs] I just want to leave you

guys with this. Uh verification is not just about trust or being trustless. Um you can actually use this technology to have serious impact within the world. Um not only like for finances but also like imagine things like your health records, right? Um you want to get access to healthare but you don't want to expose like all this information about your past um history. Um so then you're able to use uh ZK to like selectively disclose stuff to get access to medicine. Um, this can be applied to almost anything um in the world. Um, it's just about how creative you think about it. So, yeah. Uh, my name is Kofi. I'm the co-founder of Credentials. Um, feel free to contact me if you want to work, discuss, jam about this stuff.

Been researching ZK for like the last three and a half years of my life. So, I'm pretty like intuitive with how this stuff works. Um, thank you for listening as well. So, we have four minutes. if you guys some has questions in the back. But >> um so how can ZKTLS prevent fraud in the case that it's not like you have a verifiable source of income or deposits? It's it's like choosing what you want to verify. Like in the in the case of the example you gave about like the $500,000 and the uh like the lend.

>> Yeah. I can prove that okay I got a land like 500k and I can verify to you that I own 500k but like that doesn't prevent that I can misuse that and still do the fraud and is the problem with that is not verifiability is like how like the depth of layer of what you want to show the traceability of the money you got because I I got I lent this money and now I can misuse it but I can still prove to you that I own 500k and I can still conduct the fraud. But so I'm wondering like how can >> forgiven so so I guess we use that same example. Um so for instance like when the guy goes to the bank now and they like verify that he's taken out a credit

loan of 500k. Um it doesn't necessarily prevent fraud in that moment. But now you have more information to question, right? You say, "Okay, you now taken out $500,000 um with this said property in the memo. Why do you want to come and take out 500k from us with the same property?" Right? it allows you to kind of have like a layer of defense um in in that case. But like I agree with you, this isn't necessarily like going to prevent um like let's call it it's not like the the secure like magic bullet for fraud, but it's it's a layer of security, right? That that doesn't exist right now, right? we we have kind of like no insight into how stuff works, which is pretty sad because there's a lot of stuff that's digital today, but now you kind of have a way to verify things um

to then like go deeper with folks like that come into your workplace and come into your stuff um you have more insight into the data basically um to do stuff. So yeah, I I I agree with you as well. There needs to be more that needs to be done, but this is like a great first step compared to what exists today. Okay, >> thank you. Okay. So, Coffee, first off, I love how passionate you are about the stuff you're building. And when I I I I said that ZKTLS will change the world, that's that's because the builders like you or Manza, I love I love your vibe and and the good cause you're working for.

>> Uh, one question. uh you are mobile first, mobile only because you mentioned reclaim uh is there any way to use that in browser server site? >> Yeah, so um quick alpha by the way we're we're we're moving to like a strictly like on the web solution within the next couple of weeks um to prevent uh like UX barriers of having to download an app and then go through a verification process. Um, so yeah, we'll be we'll be strictly on the web like sometime on the third week in December to where now like you can now just get redirected to the Uber or like the Binance. Um, and you won't even need to see credentials or anything like being uh in the middle of

like your process. It's one of the issues that we had with like the the drivers in South Africa. They didn't want to download the app. Um, which is why like we're quickly changing that process. So yeah, thank you for asking that. >> All right. Thank you. See you later. All right. [snorts] Any more or you guys have enjoy the conversation? Yeah. All right. Thank you for listening. You can also ask questions in the hallways always. So, uh,