THE EDGE,
WITHOUT THE HYPE.
Production technique, emerging infrastructure and active research are not interchangeable labels. We keep them visible.
Rough Volatility
Very low regularity volatility models aligned with observed short-scale behaviour.
AAD
Adjoint algorithmic differentiation for many sensitivities at near-constant reverse cost.
Differentiable Pricing
Pricing systems designed for gradients across models and parameters.
Machine Learning Surrogates
Fast learned approximations to expensive pricing maps.
Deep Hedging
Learn hedging policies under frictions and non-quadratic objectives.
Neural SDEs
Stochastic differential equations with learned functional components.
Bayesian Calibration
Infer parameter distributions rather than one best-fit point.
GPU Monte Carlo
Parallel simulation and payoff evaluation on graphics processors.
Fit is not evidence.
Speed is not governance.
Experimental methods require benchmark error, stability domains, fallback models and production monitoring before they earn desk-standard status.