Projects#

Light-weight BQL market data retriever with main execution outside of BQUANT environment

Implemented a FX G10 trade blotter to display proprietary prices (pricing engine output) without markups applied. Handles arbitrage detection & core spread anomalies

Light-weight Python library to create and manage UNIX-based and Windows (service) daemon processes

Implementation of a L4 network load balancer with gateway pool manager, health checks and retries logic to retrieve various market data from a pool of LAN remote machines

ig-data helps users collecting real-time events (via LightStreamer protocol) directly from their account activities.

earnings is a lightweight and featureful Python package tailored to retrieve insightful earnings details, allowing users to collect previous and future earnings calendars, filter by US-equity ticker, define a custom portfolio and easily integrate reliable financial data into your applications.

PERSO • Techs: Python • Aug. - Sep. 2023 • GitHub

Interactive ticker lookup tool

Simple KDB+/Q wrapper of the HTTP and WS API

PERSO • Techs: Python (requests, selenium, bs4) • Jul - Aug. 2023 • GitHub

PERSO • Techs: Python (requests, selenium, bs4) • Jul - Aug. 2023 • GitHub

• Scraping, processing and storing worldwide economic calendar

• Automatic data scraping of a MS PowerBI dashboard

PERSO • Techs: Python • Jul. 2023 • GitHub

PERSO • Techs: Python • Mar. 2023 - Apr. 2023

ACADEMIC • Techs: - • Feb. 2023 - May. 2023

• Estimation of Hurst exponent from Lifted Heston model with fractional Brownian motion (Cholesky method)
• Theoretical proof of Carr-Madan formula of call option price
• EUR Call option pricer using Fourier methods for Lifted Heston model
• EUR Call option pricer using Monte-Carlo for Lifted Bergomi model
• Computation of implied volatility smiles for various equi-distant log-strikes intervals

• Statistical analysis of spread distribution, ticksize, ...
• Simulation of several point processes (Homogeneous/Non-Homogeneous Poisson, Hawkes, Cox, ...)
• Goodness-of-fit with empirical durations
• MLE for Hawkes processes calibration

• Estimate and plot the values of the estimated realized volatility when using observation frequencies ranging from 30 seconds to 15 minutes.
• Compare these estimations with the long range estimation of the volatility (based on 1 month of daily data)
• Provide some estimation of the micro-structure noise size (by using the autocorrelation between the returns at different scales)
• Plot the evolution of the estimated daily volatility of the IVE over the last year.

• Computation of the greeks (delta, gamma, theta, rho, vega)
• Implementation of the pricers using FDM (Crank-Nicholson)
• Computiation and plotting of the optimal exercise frontier

ACADEMIC • Techs: Python • Dec. 2022 - Jan. 2023 • GitHub

• Use of financial market data in order to calibrate the model
• Study of the impact of each parameter of SABR model
• Construction of implied volatility smiles

Use of ResNet 50 NN model to classify and detect blood cell types among a large dataset

• Binary classification on the level of ice quantity in Greenland
• Benchmarking several binary classifiers performances

In-depth analysis of a portfolio of diverse assets (stocks, bonds, futures/forwards, options, funds, ...)

• Studying empirically the impact of several NN architecture hyperparameters (# of layers, # of neurons/layer, activation functions, regularization (L1, L2, ElasticNet), learning rate evolution, impact of normalization + features noising)

Refactoring of the current webapp with a new, easier-to-use web application hosted on GitHub pages environment (link)

Given a set of 100 equities and fund indexes: - Preliminary data analysis and pre-processing - Global Mean-Variance - Bayesian approach - Black-Litterman - ... - Benchmark, Performance analysis & Conclusion

ACADEMIC • Techs: MATLAB, Python, R • May. 2022 - Jul. 2022

-

K-Means variation for object-distance array - Benchmark with Adjusted Rand Index computed with adj.rand.index function from package fossil - PCA with factoextra - Comparison with other algorithms - Complete example on iris dataset

Open-source TeX package to better deal with TODO tasks beacons within a thesis/long report writing in LaTeX. Check the official Git repo for more information.

Open-source Python tool dedicated to convert and export records and underlying signals from ECG data for further data analysis processing

Open-source application dedicated to analyze real-time or historical data (long, lat, alt, speed, type of motion, raw gyroscope/magnetometer/accelerometer data, pedometer, ...) In the future, ambition to embed these data processing protocols into a Grafana-like dashboard for optimized UX

Functional programing implementation of MapReduce paradigm using OCaml

Academic presentation (in French) on the numeric computation of the Fibonacci sequence - Naive and iterative approaches (Recursion & Explicit methods) - A last track : matrix writing - Application of $F_n$ • Euclid's algorithm

Academic assignment about data analysis and binary classification throughout SVM methods on breact cancer data

• Full & autonomous authentification system design and implementation
• Development of a full-featured calendar system
• Implementation of a user system with settings and dashboard features
• Deployment on production server

• Design, development, test and deployment of a data extraction/formatting software for a French real-estate agency
• Application dedicated to convert raw PDF data from huge-sized files to XLSX/JSON/CSV format using accurate computer vision solutions (OCR), implemented in Python, for advertising purposes

Development of this current website to illustrate my projects (past, present & future), my academic and technical interests and my skills

• Development of a web-scraping tool to gather statistics and historical data from various sources
• Set up of a FastAPI-dashboard for results & graphics display
• (TO DO) Statistical analysis (Pattern recognition, sequence analyzer, ...)

In the context of an internship :

• Development of document scanning request server (Python: FastAPI, ownCloud api, ...)
• Implementation of a custom authentification system, which makes a secure link for upload between a distant NextCloud session and the scanning server
• Design following the API RESTful architecture

In the context of an internship :

• Design and Development of a NextCloud application dealing with document scanning request
• Implementation of a full-featured dashboard to easily handle scanning request lifetime and status protocols
• Technical prospection for the authentification part around the following protocols : 2FA, FIDO Key, TOTP, OTP
• Handler with the authentification process using NextCloud OAuth API
• Automation of tokens validity using cron

• Implementation of a basic network speedtest tool dedicated to analyse upload/download stream speeds in order to verify the conformity of the current service in comparison with the ISP signed contract.
• After each analysis, release accurate charts ready to be plot to the end-user
• Include use of several testing servers (through the used py extension)

• Scraping & formatting of data from IMDB (MySQL db of 5 Gb)
• Implementation of a full sentiment analysis system based on NLP algorithms working on Twitter and Reddit textflows thanks to their respective APIs
• Developped a dynamic search based of several RegExp combinations on a MySQL db

• Scraping & formatting of data from IMDB
• Developped a dynamic search based of several RegExp combinations on a MySQL db

Implementation of a custom board game in C. Work subject available in the Git repo

Personal project part of an official Kaggle competition from ConocoPhillips (large oil-equipment company)

Official competition page

Development of a Qt-based application in Python to format and edit multi-pages documents and add Traffic Light Protocol rules.

Definitions & Usage

• Used functional programming paradigm and OCaml language
• Implemented generic functions for DNA strand study and genomic analysis
• Implemented general structure of phylogenetic tree

Website for the DataScIIEnce student club including : - User management systems - Showcase website for commercial and academic purposes - Ressources management center (MOOCs, Kaggle, team articles, ...)

Studied, implemented and tested several operational research algorithms dedicated to solve an adapted Shortest Path Problem, to optimize a vessel route, using random-generated and scraped data.

Created, designed & developed a small standalone robot, provided with an obstacle avoidance system using Arduino's technologies and several electronic components (C, Matlab, Python)

• Commercial study of various solutions (LiDAR, Ultrasound & Infrared)
• Use of ultrasound systems for avoidance protocols
• Development of the embedded systems
• Enhancement of a current LCD screen management C library to include the I2C bus norm
• Performance analysis on real environment with various-sized obstacles
• Full development, deployment and English/Spanish translation of the technical report as a static web app

Theoretical study and numerical implementation of several pricing methods

• Discrete-time solution using the CRR model
• Enhancement of the first solution using the Black-Scholes method (Monte-Carlo simulation and Closed form)
• Stochastic study of the convergence between discrete and continuous models
• Numerical computation of the option’s put price using Finite Difference methods (explicit, implicit and Crank-Nicholson stencils) on Black-Scholes PDE’s formula

• Study of European option in financial engineering litterature
• Transformation of Black-Scholes-Merton PDE's equation into diffusion equation problem
• Implementation of C++ pricing tools using strong and optimized OOP architecture
• Development of an enhanced and complete wrapper of SDL (Smart DirectMedia Layer) libraries to build professional scatter plots from scratch