Projects

Objective

Design a private, adaptive conversational AI system for education, focusing on data-driven personalization, process automation, and institutional data sovereignty.

Key Contributions

• Designed and structured an internal educational database to centralize academic resources, student interactions, and curriculum-related data, enabling efficient querying and downstream analytical use.
• Developed a workflow-based automation layer for structured query processing, content filtering, and dynamic recommendation of learning resources based on user interactions.
• Evaluated architectural trade-offs between cloud-based and local deployments, assessing the replacement of external LLMs with locally hosted open-source models (LLaMA 3, Mistral) to ensure privacy and institutional control.

Objective

Develop a digital twin of a microscope to simulate its behavior and support educational and demonstrative use, in partnership with the LEM3 laboratory (Laboratory of Microstructure Studies and Mechanics of Materials).

Key Contributions

• Collaborated with industrial partners to understand the physical principles, functional requirements, and mechanical architecture of microscope systems.
• Developed a digital twin integrating 3D modeling and dynamic simulations to replicate component behavior and system interactions.
• Created simulation-based usage scenarios to illustrate real operating conditions and enhance pedagogical applications.

Objective

Develop a Python-based simulation game inspired by poker to model probabilistic decision-making and analyze outcomes using structured data.

Key Contributions

• Built an interactive game interface enabling real-time gameplay and visualization of decision scenarios.
• Designed the software architecture using object-oriented programming principles (classes, inheritance, encapsulation).
• Implemented structured data processing to analyze hand distributions and develop statistical analysis modules for probability evaluation.

Objective

Design, build, and analyze a scaled electromagnetic aircraft launch system (EMALS) to model and optimize launch dynamics under controlled conditions — spanning physics modeling, hardware construction, and experimental data analysis over two years.

Key Contributions

• Developed and validated a physics-based model of an electromagnetic aircraft launcher to predict velocity and performance, using experimental data for iterative calibration.
• Integrated position sensors and developed custom algorithms to derive real-time velocity and acceleration data, enabling the extraction of key flight parameters.
• Designed and built a scale electromagnetic launcher (aluminum frame, coils, power electronics, motion sensors) enabling repeatable high-speed tests for performance benchmarking and system optimization.