Independent Researcher
Dec 2025 — Present India
Agentic AI workflows and LLM applications
Following graduation, I pursued independent research focusing on agentic AI workflows and the application of LLMs to both agent-based models and conceptual modeling (example project).
Toulouse, France
Fine-tuning of Vision-Language Models on Satellite Images
Under the supervision of Fahd Benatia, I conducted research on the underlying AI model of Airbus Geo Explore. The work involved reviewing state-of-the-art fine-tuning techniques for VLMs in remote sensing, building a benchmark suite to compare models and training strategies, and implementing end-to-end fine-tuning pipelines on GCP.
Research Intern (LLMs & Simulation) — VMASC
Aug 2024 Suffolk, VA, USA
Applied LLMs to simulation and decision-making
Advised by Philippe J. Giabbanelli, my research focused on making complex Agent-Based Models interpretable for non-technical decision-makers. I built an end-to-end framework that parses NetLogo simulation metrics, leverages advanced prompt engineering with GPT-4o and Claude 3.5 Sonnet to analyze data trends, and condenses the text using both extractive and abstractive summarization. The resulting explanations were validated through comprehensive automated scoring and structured human evaluations.
Research Intern (LLMs) — Miami University
May 2024 — Jul 2024 Oxford, OH, USA
LLMs for conceptual modeling and multimodal evaluation
I investigated how LLMs can be integrated with simulation and conceptual modeling. Philippe Giabbanelli was my supervisor. My contributions included testing LLMs on multimodal graduate-level exams using both text and slide decks, applying a Design of Experiments methodology to merge complex causal maps using LLMs, and utilizing advanced prompt engineering and style transfer to translate agent-based models into empathetic stories.
AI Research Engineer Intern — EuroMov
Jan 2024 — May 2024 Alès, Occitanie, France
Generative AI applied to artistic movement analysis
Supervised by Gérard Dray, my research stood at the intersection of AI and art by generating digital artwork from human movement data. I processed raw sensor spreadsheets from an artist's movements and used a remote GPU container to run StableDiffusion with ControlNet and specialized models. I developed an iterative workflow that combined image generation with MATLAB scripts for depth-based image composition, accepting uncertainty to create novel visual compositions and study the expressive potential of diffusion models.
Data Analyst Intern — Centre Hospitalier de Thuir
Nov 2023 — Dec 2023 Thuir, Occitanie, France
BI dashboards to modernize hospital data reporting
I modernized hospital data reporting under the supervision of Clément Naudy, by replacing manual export processes with real-time QlikView dashboards connected to administrative and medical ERPs. I wrote SQL queries and QlikView load scripts to extract, transform, and model hospital data, building dashboards to track key indicators such as medical activity or absenteeism. Finally, I authored technical documentation to ensure project sustainability.
Market Analyst Intern — Haï Mask
May 2023 — Jun 2023 Alès, Occitanie, France
Market study for a connected sports mask project
Supervised by Maxime Menardo, I conducted a 5-week market study that included a PESTEL analysis, competitive mapping, and target audience profiling. I surveyed 22 high-level athletes to validate product features and test price sensitivity, ultimately building a 4P marketing plan, a CRM strategy, and a go-to-market action plan. I also designed a multi-platform communication strategy with promotional content mockups and analyzed pro sports club budgets to identify B2B prospects.