Collaborating with enterprise organizations to design and implement quantitative decision-making platforms, leveraging rigorous data analytics and fault-tolerant distributed systems to solve mission-critical business challenges at scale.

Led the evolution of cybersecurity through artificial intelligence and advanced risk quantification at Balbix.

As a key member of the Cyber Risk Quantification (CRQ) group within Balbix’s AI Team, I transformed complex cybersecurity data into actionable financial insights that bridge the gap between technical teams and executive decision-makers.

Key Achievements:
• Architected and implemented an innovative ‘Next Best Steps’ feature that revolutionizes how organizations prioritize cybersecurity initiatives. By leveraging Game Theory’s Shapley values, I developed a sophisticated pipeline that quantifies the financial impact of individual vulnerabilities across enterprise networks. This empowers IT teams with clear, high-impact action items while providing executives with precise risk visibility. Technologies: Apache Spark, Apache Airflow, Delta Lake, FastAPI, PostgreSQL.
• Spearheaded the redesign of Balbix’s core asset valuation engine following the launch of Balbix D3. Created a unified mathematical model that dynamically assigns monetary values to all asset types (devices, applications, users) based on customer-specific parameters. To optimize performance and reduce operational costs, I engineered an efficient pipeline using Delta-RS and DuckDB. This foundation powers all CRQ capabilities across the platform. Technologies: Delta-RS, DuckDB, Apache Airflow, Delta Lake, FastAPI, PostgreSQL.
• Created a Loss Exceedance Curve analytics tool that estimates the probability of surpassing defined financial loss thresholds over a year across diverse cyber attack scenarios. Leveraging advanced mathematical modeling and optimized computation techniques, this solution integrates multiple risk factors to deliver real-time, actionable insights tailored for each customer and asset group. The framework seamlessly works with our asset valuation process, enabling clear, customized risk assessments. Technologies: Python, Airflow, FastAPI, Numpy, Pandas, and PostgreSQL.
• Optimized Balbix’s cloud infrastructure to reduce cost of Apache Spark services by 50%.

During the Summer 2023, I perfomed an internship at Balbix within the AI/ML Team to help them improve their modelization of risk’s over a network and develop a new scoring metric for prioritizing vulnerabilities' resolution.

I continued my work at Balbix during the Fall 2023 semester as a part-time intern while finishing my studies at Columbia University. I worked on improving Balbix’s CRQ model using new datasets and state-of-the-art modeling techniques.

Key Achievements:
• Architecting machine learning models using graph theory and MITRE’s CAPEC dataset to enhance Balbix’s Cyber Risk Quantification (CRQ) platform, which quantifies cyber risk exposure and potential financial impact in business terms ($, €, £, ¥).
• Pioneered a novel cyber risk scoring methodology using Shapley values to help clients intelligently prioritize 1000s of vulnerabilities' mitigations based on potential financial impact. The approach considers each vulnerability’s marginal contribution to overall network risk, providing more nuanced insights than traditional CVSS scores.
• Optimized scoring algorithm runtime by 500%, compared to base implementation, by parallelizing computations in Apache Spark using PandasUDF, enabling hourly updates of risk scoring of millions of devices and vulnerabilities.
• Developed, using Dynamic Programming, a risk propagation model on a network of assets leveraging MITRE ATT&CK Framework adding additional data to my previous work.
• Created a cybersecurity handcrafted dataset to test my model and verify our assumptions.

Teaching Assistant for the Department of Mathematics at Columbia University for Calculus II (Fall 2022 & Spring 2023), for Linear Algebra & Probabilities (Summer 2023) and for Calculus I (Fall 2023).

Key Achievements:
• Evaluated and overlooked assignments, projects and exams performance of over 350 students.
• Enhanced students learning by providing individualized assistance to follow lesson plan.
• Top 9 finalist of the University-wide ‘2023 Presidential Awards for Outstanding Teaching by a Graduate Student Instructor’.

Research Assistant in Jacobs lab: Electrophysiology, Memory, and Navigation Laboratory (https://jacobslab.bme.columbia.edu/)

Worked on a research project to develop classification machine learning models (using Linear Classifier, SVM, and MLP) to predict visual stimulus using electrocorticography neuronal signals with statistical significance.

• Developed for a startup machine learning models using Linear Regression, SVM, or XGBoost to predict time-series of daily attendance levels over a week of restaurants in Paris.
• Automated the collection and the pre-processing of open data from 6 different sources (weather, road traffic, events, etc.) depending on the location of the restaurants.
• Released a statistical method using Meta Prophet framework to predict attendance levels with 95% confidence intervals.
• Compiled a report to explain the functioning of my code and the impact of different parameters.

• Organized conferences with prestigious alumni (Paul-François Fournier, Michel Combes, Fred Potter, Jean Schmitt).
• Expanded corporate relations (partnerships with PwC, Strategy& and AlumnEye).
• Handled the communication of the association (Facebook, Instagram, LinkedIn and the website).
• Managed all the association’s administrative procedures and archives.
• Realized a conference on personal finance to introduce this subject to Télécom Paris' engineering students and explain them why they should invest as soon as possible, where they could start investing in and how they could develop their portfolio in the future
• Participated, as Secretary-General, in the definition of the association’s strategy to make it grow.

• Tutored 4 students from 6th grade to 12th grade in mathematics, physics, French, and chemistry for preparation to French examinations.
• Helped students with post-baccalaureate education counseling.

Managed a team of 40 people to organize Télécom Paris' 2021 career fair
Results : more than 500 students, 78 companies, €200k turnover, company satisfaction: 4.1/5 and student satisfaction: 4.4/5.
Handled:
• corporate relations (prospecting, customer service, etc.);
• logistics (choice of furniture, organization before the event, etc.);
• communication (administrator of Facebook & Instagram pages);
• the website (development of a digital business card solution via a QR Code system, CV library, interactive map, etc.);
• the organization of the association’s events (association campaigns, student parties, distribution of €55k for student projects, etc.).
As an elected member of the association’s board to represent both students' and school’s interests, I was always in discussion with the administration and corporate relations of Télécom Paris. I was also responsible for major decisions concerning the association.

Joined a team at Sopra Banking Software as part of the Undergraduate Program set up by Sopra Steria.

Key Achievements:
• Automated their internal documentation website (Python & GitLab) for referencing solutions deployed for their customers;
• Automated verification of bank security certificates from providers (Python & AWS Lambda) to send the necessary information to the Support teams.

I also learned more about the group by discovering all the verticals and subsidiaries of Sopra Steria. I was able to spend a full day with the HR manager and member of the Comex of Sopra Steria Group: Jean-Charles Tarlier.