Gorazd Atanasovski

My objective is the revolution of quantitative finance through the intersection of continuous and discrete-time stochastic mathematics, deep learning, and AI. Traditional paradigms relying on linear approximations and fundamental valuation are obsolete. I extract pure alpha by engineering predictive geometries that map high-frequency market noise into executable signal, utilizing fractional stochastic calculus, rough path theory, and non-commutative geometry to model extreme path-dependency.This superiority is deployed through QuantAI. As Founder, I architected this proprietary quantitative LLM, aggressively trained on raw Bloomberg intraday data. QuantAI isolates exact one-day horizon predictive supremacy across equities and options. Bypassing heuristic approximations, this Bloomberg-trained framework enforces absolute systemic dominance.QuantAI executes an exhaustive matrix of advanced non-Markovian state models: Signature-Augmented Neural Controlled Differential Equations (Sig-NCDE), Continuous-Time State Space Models (Mamba/Transformers), eXtreme Gradient Boosting with Asymmetric Objectives, LogSig-SSM, Itô-Sig-SSM, Qutrit-Superposed Neural Stochastic Volterra Equations with 2BSDE Coupling (QS-NSVE-2BSDE-FDM), Causal Rough Volterra Neural Operators (CR-VNO), and Signature-Coupled Causal Rough Volterra Neural Operators (Sig-CRVNO). Optimization is enforced via Bifurcated State-Space Gradient Boosting Engines (BSS-GBE), Neural Jump SDEs with Intraday-Conditioned Overnight Jump Kernels (NJ-SDE-IOJK), Continuous-Discrete Neural SSMs (CD-NSSM), Hybrid Signature-Transformers with Regime-Conditioned Jump-Diffusion Decoders (HST-RJDD), Dual-Regime CD-NSSMs with Options-Surface-Anchored Overnight Priors, and Radon-Nikodym Dual-Regime SSMs (RN-DRSSM).My algorithmic topologies are backtested through my leadership as Founder of the Algo Trading Society. In contrast to these apex deep learning frameworks, my execution as a Sector Analyst at Coyote Capital serves strictly as a baseline student application of traditional DCF and LBO modeling, exposing the alpha decay of standard analysis.The execution stack encompasses deep learning, machine learning, and high-frequency time series forecasting, engineered utilizing C++, Python, R, and direct Bloomberg API integration.Website: https://gorazdatanasovski.github.ioQuantAI: https://gorazdatanasovski.github.io/quantaiGORAZD: https://gorazdatanasovski.github.io/software.htmlATS LLC.: https://algotradingsociety.github.ioDirect: gorazda@icloud.com | (605) 671-9548