Alexandre Le Borgne

2026 – LLM runtime verification: design of guardrails for LLMs ensuring the correctness of responses via formal methods, implementation of AI workflows with LangChain (Python, VSCode, Agentic AI, LangChain framework).

2025 – Machine learning for security: improvement of attack detection on virtualized environments via an anomaly detection approach targeting container breakouts (model design, training, dataset preparation in Python/PyCharm).

2025 – Simulation for cloud computing: design and development of simulators to evaluate memory pooling and swapping strategies at cluster scale, scenario generation, load randomization, genetic algorithm to explore the parameter space, support for C/C++ development of the swapping simulator, with product line reward.

2024–2025 – Code generation for network equipment: design of a Datacom configuration DSL to generate product code directly from specifications, tool delivered to the department (Eclipse, Xtext, EMF).

2023 – Coverage code generation: development of a Software‑in‑the‑Loop coverage tool for the automotive hardware code generation platform, enabling measurement of test harness coverage in a back‑to‑back approach (Java, integration into commercial delivery).

2022–2024 – Requirements formalization: leading an academic collaboration with the Aristotle University of Thessaloniki to define an intermediate specification language that automatically generates formal properties, then exploring the use of LLMs to transform natural language into this intermediate language (DSL/DSML, M2T transformations with Eclipse, Xtext, EMF), used in defining requirements for the verification of data storage scenarios and network architectures.

Development of mixed reality human-computer interfaces for manipulating Smart City data.

Design of model-based approaches to enable citizens and decision-makers to create their own applications within the urban ecosystem.

Proposal of an IoT approach providing open access to real-time environmental data.