Designing a signature scheme that survives quantum computers

At DRDO's Centre for AI & Robotics, I got to work on a genuinely intimidating problem: digital signatures that stay safe even against a quantum computer. Today's signatures lean on math that a big enough quantum machine could crack. So a bunch of us are racing to build ones it can't. I designed a scheme we called CMSS-PPR. Without drowning you in algebra: multivariate schemes like this can be broken by clever attacks with names like MinRank and Gröbner-basis. They exploit hidden structure in the equations. The core idea I contributed is something we named Non-Affine Ring-Permutation Layers. Stacking these layers pushes up the "regularity degree" of the system — which, in plain terms, means the shortcuts those attacks rely on stop working. The structure they'd normally grab onto isn't there anymore. I implemented the whole thing in Python — key generation, signing, verifying — working across all five NIST security levels. And on paper, I reduced its security down to a problem that's known to be hard, showing it holds up better than older schemes like Rainbow, UOV, and HFE. It's under review at a journal now, and the DRDO scientists were kind about it. The lesson that stuck: in cryptography, the attack teaches you the defense. You can't protect against something you haven't tried to break.