Most of my time goes into two things: shaving latency off derivatives risk engines at Sberbank CIB, and building Linux software that talks straight to the hardware.
By day, I work on C++ adapters and calculation pipelines for Rates, FX, and Equity derivatives. We are moving pricing and risk out of legacy Murex monoliths into an in-house calculation service. That means zero-copy data mapping, streaming market data, Greeks injection, and Monte Carlo scenarios for XVA.
On weekends, I write Wayland daemons in C23, benchmark SIMD routines, and train neural operators on GPUs.
- Work: Middle Quant Developer at Sberbank CIB (Global Markets / ДГР)
- Education: MIPT ФПМИ — BSc in CS ('25), MSc in Corporate Strategy & Quant Finance ('26), with ML coursework from Yandex School of Data Analysis (ШАД)
- Setup: Arch Linux, Hyprland, pure terminal workflow, Neovim
derivatives — GPU rough volatility calibration and structured derivatives engine
An institutional quantitative derivatives platform built for my master's thesis at MIPT, covering a 16-model zoo from rough volatility to multi-asset structured products. Features a 3-tier hybrid Multi-Asset EQAutocallable Engine (+1.05 bps mean error vs Murex MonteCarloWithAAD, 0 ULP diff on 36 deterministic cashflows, 1.40 ms warm latency for PV + all 8 Greeks, and 67,114 trades/s batched GPU throughput), a Triangulated FX & Commodity Cross Engine (0.35 bps RMSE, 0.196 ms p50 latency, 13 Taylor-jet Greeks), and a FiLM-conditioned Fourier Neural Operator (FNO) pricing rough Heston surfaces in <1 ms (
wwal — High-performance Wayland wallpaper & dynamic theming ecosystem (v1.2.0)
WayWal replaces the fragmented desktop stack (swww/awww, mpvpaper, linux-wallpaperengine, glpaper, wpaperd, and matugen/pywal) with a single 1.66 MB C23 daemon (wwald) and client (wwal). Decodes 4K video via hardware VA-API straight into DRM PRIME buffers (~0.3% CPU), runs native Steam Wallpaper Engine Workshop scenes without Wine/Proton, extracts dark/light Material You palettes and generates system color configs in 0.49 ms end-to-end, and shifts to 0.00% CPU/GPU with zero wakeups whenever windows cover the screen. Features 10 GPU transitions (1,668–2,734 FPS at 4K), real-time PipeWire audio-reactive shaders, sub-millisecond picker (wwal pick in 0.93 ms), and official packages across AUR, Fedora/openSUSE Copr, Debian .deb, and Nix.
Fast-hash-table — AVX2 SIMD hash table
A closed-addressing hash table in C aligned to hardware cache lines. Uses 256-bit AVX2 intrinsics (_mm256_cmpeq_epi8) to compare 32 keys in parallel in a single cycle, with hardware CRC32 (_mm_crc32_u64).
aurora-greeter — Hardware-adaptive SDDM login theme
An SDDM theme for Arch Linux with video backdrops, packaged on the AUR. Includes an in-greeter settings drawer (Alt+S) so you never have to edit config files by hand, plus automatic 4K → 2K downscaling on smaller screens.
HyprGlyph — Hyprland keybinding parser
Parses bindings directly from Hyprland configuration files into structured JSON for desktop overlay drawers, avoiding manual cheatsheet tables that rot when window manager rules change. Written in Python with native support for both .conf and Lua configs, backed by 80 passed tests.
Execore_Language — Statically typed language engine in ISO C++23
An interpreted language runtime combining Python-style indentation syntax with explicit static types (int, float, char, str, list), first-class lexical closures, and runtime method dispatch. Engineered for cache locality and zero-allocation execution paths: tokens stream through Structure-of-Arrays buffers (TokenBufferSoA) for 100% L1 cache density, AST nodes allocate from PMR monotonic memory arenas (std::pmr), and errors propagate via C++23 std::expected monadic chaining. Tagged std::variant primitives evaluate without heap hits. Passes a 5-tier test matrix (unit, BDD, 2,000 property tests, LibFuzzer, mutation testing) and achieves >100 MiB/s lexer throughput (~4.8M tokens/s) under PGO and ThinLTO.
I maintain custom, low-overhead environments for high-refresh Linux desktops:
- shell (C++20 / Qt6 / QML): High-performance desktop shell fork engineered for minimal latency and footprint: -60% VRAM, -52% RAM, zero idle wakeups,
RTLD_LOCALdlopendynamic library isolation, 64 MiB LRU texture cache, and SDF shader culling. Includes smart dynamic application grouping (Super+D), visual X11 app quarantine, and interactive region OCR. - caelestia (Hyprland / Lua): Desktop dotfiles with native Hyprland 0.56+ Lua configuration rules, AwesomeWM-style independent multi-monitor workspace group emulation (
wsaction.fish), and deepwwalddaemon integration with live animated previews.
- Systems & Low-Latency: Modern C++ (17/20/23), Modern C (C23), x86-64 NASM Assembly, AVX2 SIMD, Linux
io_uring, DMA-BUF, DRM/GBM, Wayland protocols, POSIX. - Quant & ML: Rough volatility, Black-Scholes, Greeks (Δ, Γ, ν), XVA (CVA, PFE), Fourier Neural Operators (FNO), Equivariant GNOs, Multi-Fidelity PDE Ensembles, PyTorch, CUDA, TensorRT, Murex FLEX API, Lean 4.
- Environment: Arch Linux, Hyprland, Meson, CMake, Git, Docker, Go, Python.
"Premature optimization is the root of all evil — unless you are out of clock cycles before the market moves."
Telegram: @execorn · n.biryukov.04@gmail.com · Moscow, Russia



