Pramaana

Derives pseudonymous credentials from identity data through hardened commitments and a document-to-face enrollment check.

Key takeaways

01

A reusable personhood credential separates proving uniqueness from repeatedly disclosing identity documents.

02

One enrollment derives a master key that can generate separate pseudonyms for different platforms.

03

Near-infrared face matching checks whether the enrollee matches the face on the submitted ID.

Summary

Pramaana is motivated by the tension between proving eligibility or personhood and repeatedly disclosing identity documents. The presenter framed the system as storing no conventional database and as targeting population-scale, post-quantum personhood, but those security and scale properties were not independently established.

The presented key-hardening pipeline begins with a 36-bit secret derived from identity information, then applies Argon2id stretching, a blinded oblivious pseudorandom function, and rate limiting. A PII-anchored lattice commitment is described as the next stage; its deterministic, one-way, and quantum-safe properties remain speaker claims rather than audited findings.

During the described enrollment flow, personal details are used as an entropy source and hashed before a commitment is registered on-chain. Verification then creates a master Phi key from which the user can derive separate pseudonyms for different platforms.

The biometric step extracts identifiers from an uploaded document and compares the user's face with the document image. The presenter proposed employee verification, background checks, financial KYC, age checks, and digital public infrastructure as use cases, but the talk did not validate anti-Sybil effectiveness, deployment scale, or integration readiness.

How it works

01

Match the person to the document

Enrollment extracts identifiers and a face image from an identity document, then uses near-infrared recognition to compare the user with the document.

02

Harden the source secret

An identity-derived 36-bit secret passes through Argon2id stretching, a blinded OPRF, and rate limiting.

03

Register the commitment

Hashed personal details feed a PII-anchored lattice commitment that is registered on-chain.

04

Derive platform pseudonyms

The resulting master Phi key generates separate pseudonyms for different platforms.

Transcript

So my project is called Pramana. The core ideology behind Pramana is to prove that you're real without storing anything. This is a post-quantum proof of unique personhood built for identity system at a population scale. Let us meet Bob. Until now, we have discussed about Alice, right? So we have a new character called Bob, and let's explore what's happening within his life. Bob knows he's being watched. Every time he buys shoes, every search engine, every literal shopping mall that he goes to, every Instagram page that he visits to, he's constantly bombarded with ads of buying shoes. So he just knows that his phones are tracking him, every search history is tracking him, and he's getting a bit paranoid.
Bob literally hates giving his passport when he needs to open his bank account or even any account when he's is doing an international travel. When he has to rent something, when he has to buy something, he has to give his personal identity information, and he's really worried about it. Bob is really worried about his identity because there is a constant news of all these millions of records getting stolen. Why even worry about records getting stolen? How does it matter to Bob, right? A random Dan can actually open a credit card on Bob's name and literally spend his money while Bob dies in credit debt. So what can Bob do, right? So also Bob realizes the 12 digits that are printed on his card, any identifying system, when you show it to any random stranger or a CVS pharmacist, he just knows that he's giving his personal data.
But until now, cryptographers told him, "Yeah, this data is pretty secure. Nobody's gonna see it." But then comes the quantum computer in the picture. It is extremely guessable, and now the fact that you have literally shared your information all the way in the web, your data is practically outside. So the key point here is the safest database is when you have no database. Be like Bob. Don't give your passport when you need to get that room in a Paris vacation. Use Pramana, the authentication system that stores nothing, and be like Bob. And you are Bob. There are many Bobs in this room. What is the problem that I'm exactly trying to solve? All the world's biggest ID systems, including KYC Aadhaar in India, which is being used by three million people right now, is centralized by design, and it relies on classical cryptography and has no relationship with Quantum Horizon.
All the stored biometrics are great, but after a point when you have to reset your password, your biometrics also become a liability. A national ID number is not a secret. You're literally giving that to a CVS cashier, so this is not gonna be a secret that you are gonna hide from everyone, right? So how do we take this, randomized digit on your national ID and make it into something more robust, right? So here, based on the IC3 paper, Anonymous Self-Credentials, here we have a thirty-six bit secret, and after that we go through three layers of additional encryption using Argon2 id stretch, blinded OPRF, and rate limiting so that we can get a final key that is much more robust that people can actually use.
And this is the PALC commitment pipeline that is inspired from Anonymous Self-Credentials paper, and this is a PII-anchored lattice commitment, which is deterministic, quantum safe, and one-way. So for the product itself, here you give your PII and your personal details like date of birth and everything. Don't be worried here because this is gonna be used as your proof of enrollment. And after you're enrolled, it just takes this information and uses, it as a source of entropy, and all this information is then get hashed. And then all this, information is used for the PALC enrollment, and you register it on the chain. And finally, you verify and create this master identity called as the Phi key, which you can literally use to create multiple pseudonyms across different platforms, which you can use it for different purposes.
And you can also not do Sybil attacks because it also comes with a biometric and human verification system. More about it. And additionally, if you can check, you can also check whether your quantum, wallet is like quantum safe or not, and it's based on the public key exposure. And finally, all your records are stored in a dashboard where you can see all your different transactions, whatever you have done, and it's completely open source, and I believe in open source, so anybody can take it and use it as a source of authentication in your system. And the biometric system that I was talking about is actually more robust than what Apple ID uses right now. So this is an NIR facial recognition system.
So what it does is that when you upload your passport or your document, it actually assumes that the document that you're uploading is legit, like TSA, right? So it takes that particular document and then it scrapes all the different information and identifiers in the document and then matches your face with the face on the document itself. Unlike the different, facial ID systems that you see in the world, which kind of stores your face so that it can fetch this biometric system back from you, it kind of compares your face with the face on your personal document so that you know that this is the person who is using this particular, you know, credential or an authentication system.
So nobody can take it, not your friends. Nobody can, like, share it. And what is the different use cases for it? Firstly, it's used for digital public infrastructure, especially for employee verification and background check, and it is highly scalable, and it can verify you without any surveillance. It can be integrated with fintech KYC and age checks. It's quantum safe from day one, and it fits into existing systems and not rips out the existing systems. And one of the biggest collaborators I have currently is the Amravati Quantum Valley, which is an initiative taken up by the state government of India. And currently, we are also trying for a collaboration with the central government, and this currently runs on the biggest population that is India.
And there are three different tracks. That is the quantum emergency, privacy regulation, and proof of personhood. And this is about me. You can check me on LinkedIn. I'm Vaishnavi Rasika. I won't, go more on this. And to PiCreds, I have tried to integrate you onto my platform, especially using the NTT and the LLMs, within my pipeline. So if you guys are here, I would love to collaborate with you, do more research, and make this project better. And the main idea is that a billion people shouldn't have to be stored just so that they can be counted.