∫
Quantitative finance
Backtesting engines, pricing, Greeks, VaR and regime-aware allocation - from Itô to production C++.
dSₜ = μSₜdt + σSₜdWₜ
Hi, my name is
Ibrahim Laklaa.
I turn mathematics into working code.
Quant developer in Paris. I build pricing and risk tools, train models on noisy financial data, and spend my evenings with quantum circuits.
01.About me
I live at the crossroads of applied mathematics, artificial intelligence and quantum computing - turning proofs into simulations, and simulations into tools people can actually run.
I came in through competitive math prep (MP*), then drifted to where mathematics comes alive: PDEs, stochastic processes, neural differential equations and, more and more, Hilbert spaces and quantum algorithms.
Today I build quantitative tooling at Louvre Banque Privée and co-run NotitiaPlatform, an AI-augmented newsroom. Off-screen: chess, arXiv, and long runs.
Tools I use:
- ▹Python
- ▹C++
- ▹Julia
- ▹Rust
- ▹Kotlin
- ▹Java
- ▹TypeScript
- ▹Next.js
- ▹React
- ▹Vue
- ▹PyTorch
- ▹JAX
- ▹TensorFlow
- ▹scikit-learn
- ▹Pandas
- ▹Qiskit
- ▹PennyLane
- ▹LaTeX
- ▹Manim
- ▹PostgreSQL
- ▹Docker
- ▹LLM Agents
- ArabicNative
- FrenchBilingual
- EnglishProfessional
What I work on:
λ
Machine learning & agents
Deep learning on noisy time series, explainable models, and agentic LLM pipelines with an audit trail.
∇θ L = E[∇θ log πθ · A]
|ψ⟩
Quantum information
Hilbert spaces, density matrices and circuits - Qiskit, PennyLane, and a deep curiosity for what computes.
S(ρ) = −Tr(ρ log ρ)
02.Where I've worked
- C++PythonJavaKotlin
- 01Build backtesting and strategy-simulation tooling in C++ and Python across equity, rates and FX desks.
- 02Implement factor-analysis models and risk indicators (VaR, stress scenarios, Greeks) in Python and Java.
- 03Engineer a multi-asset research engine: simplified pricing, Greeks computation and dynamic allocation.
- 04Ship interactive dashboards and data automation in Python / Kotlin for PMs and risk teams.
03.Things I've built
Other projects
2022
Image Classification with CNN
Python · Keras · TensorFlowCodeA convolutional neural network for image classification, built with Keras on TensorFlow.
2022
Hand Gesture Recognition
Python · OpenCV · Deep LearningCodeHand gesture recognition using deep learning and OpenCV.
2021
Birth Chrono
JavaScript · HTML · CSSCodeLive Birthday wisher with a countdown and a shareable generated link.
04.Research notes
α
WP-2026-001quantml
AlphaSeek - Regime-Aware Portfolio Allocation
Abstract
Detects latent market regimes via HMM, GMM and spectral clustering, engineers 13 market features, and adapts allocation across five methods. Walk-forward backtests (2009-2023) show regime-conditional strategies beat naive baselines. 9-page LaTeX paper with auto-generated charts.
C
WP-2026-002quantmlxai
Credit Risk & Options Pricing - A Comparative Study
Abstract
Five classifiers (logistic regression, random forest, gradient boosting, XGBoost, neural networks) on synthetic credit data, side-by-side with three European call pricing methods: Black-Scholes analytical, Monte Carlo, and a neural-network approximator. Explainable-AI lens throughout.
05.Little lab
Three small tools I like to play with. Everything runs in your browser.
Price
6.927
Δ
0.569
Γ
0.0253
ν
0.278
Θ/d
-0.021
ρ
0.250
d₁ = 0.174 · d₂ = 0.019 - payoff at expiry (grey) vs present value (blue)
06.How a thought travels
What I model in machine learning, the brain does on 20 watts. Here is how an image becomes an idea, step by step.
step 1 / 6
Light becomes electricity
A photon hits a photoreceptor in the retina. A single molecule, retinal, changes shape, and a chemical cascade turns light into a change of voltage. The world is now a pattern of electrical signals.
~120 million rods per retina
86 bn
neurons in a human brain
~10¹⁴
synapses, give or take
~20 W
the power of a light bulb
120 m/s
fastest signals on myelinated axons
Orders of magnitude from the literature (Azevedo et al. 2009 for the neuron count). The membrane curve model: Hodgkin and Huxley, 1952.
07.Get in touch
Open to research collaborations and quant / ML roles. My inbox is always open, I try to answer everyone.