Mehdi Makni

Third-year graduate student at the Operations Research Center at MIT; quantitative research intern at Citadel (Data Strategies Group, alternative data research). Background in mathematics and computer science from Ecole Polytechnique, with research work in statistical machine learning and uncertainty quantification.

Notes8

• Quant employment verification: CONFIRMED. Actual, dated roles exist beyond student programs: Quantitative Research Intern at Citadel, Data Strategies Group (New York, Jun-Aug 2026) and Quantitative Analyst Intern at ENGIE Global Energy Management & Sales (Paris, Jul-Aug 2022), both stated on his MIT-hosted academic CV and corroborated by his LinkedIn. Employer classification: Citadel (citadel.com) is a global hedge fund / market maker whose Data Strategies Group does alternative-data quant research; ENGIE GEMS is the energy-trading and asset-management arm of ENGIE. Identity is supported: the Citadel internship and the same name, MIT ORC affiliation, advisor and career history appear consistently across the CV, LinkedIn headline ('MIT PhD | Quant research intern @ Citadel'), MIT ORC directory and a Citadel-team aggregator. Note the 2026 Citadel internship dates come from his CV; the LinkedIn-derived index may still show the internship as current after it ended Aug 2026.Aug 2026
• PhD candidate at MIT Operations Research Center (Sep 2023-May 2028), advised by Prof. Rahul Mazumder; GPA 5.0/5.0 per his own academic CV.Jan 2026
• Public academic email [contact omitted] appears on both his own CV and the MIT Operations Research Center student directory.Jan 2026
• Interned as Quantitative Analyst at ENGIE Global Energy Management & Sales, Paris (Jul-Aug 2022), building statistical software linking TTF natural-gas forward-curve moves to fundamental supply-demand factors; ENGIE GEMS is the group's energy trading/asset-management arm, so this is an energy-trading quant role. Corroborated by his CV and LinkedIn.Jul 2022
• Valedictorian of Ecole Polytechnique's BSc in Mathematics and Computer Science (Sep 2019-Jun 2022, GPA 4.24/4.3); a Tunisian community post and his CV both state the top-of-class honor.Jun 2022
• Co-authored five research papers 2023-2025: 3BASiL and TSENOR (both NeurIPS 2025), Differentially Private Sparse Fine-Tuning (KDD 2025), a Unified Sparse+Low-Rank Decomposition framework (CPAL 2025, oral), and Conformal Prediction for Federated UQ (ICML 2023).2025
• Represented Tunisia at IMO 2018, earning an Honourable Mention (14 points, rank 307 / 48.5%), with a top score on P1, per the official IMO results table.Jul 2018
• Research focus is LLM compression (pruning, quantization, low-rank decomposition), parameter-efficient fine-tuning, data valuation (influence functions), differential privacy (sparse DP-SGD), and GPU-accelerated optimization, per his own site.2026
• Advisor is Prof. Rahul Mazumder at MIT; earlier advisors were Prof. Eric Moulines and Prof. Merouane Debbah at Ecole Polytechnique - a warm path into the statistical-optimization and ML community.2026
• Served as Applied Scientist Intern at Amazon, Seattle (May-Aug 2025), building a multi-modal framework for Amazon Pharmacy in the Gen AI / LLM group.May 2025
• Worked as Research Engineer, then Research Intern, at Huawei's Lagrange Mathematics and Computing Research Center in Paris (2022-2023), supervised by Prof. Eric Moulines and Prof. Merouane Debbah; work on federated-learning uncertainty quantification led to an ICML 2023 paper.2022
• Won Second Prize at IMC 2021 with 25 points competing for Ecole Polytechnique, per the host university's official news page.Aug 2021