Mathieu Astruc

Airbus@airbusApplied AI EngineerCurrent

• Built a hybrid RAG system over 10k+ export licenses and regulatory documents, routing queries between a LangChain/FAISS retrieval pipeline and a NL-to-SQL path for structured data, guaranteeing zero generative approximation on field-level queries.
• Engineered the full pipeline end-to-end: chunking, embedding, reranking, token-cost optimization, an evaluation framework and a user-feedback loop, all designed for reliability in a safety-critical environment.
• Developed AI proofs of concept with cross-functional teams and presented the system to 100+ employees, driving enterprise AI adoption.

NTNU@ntnuResearch Engineer Intern

• Lead author of a paper accepted at HCI International 2026 (Montréal) on a real-time computer-vision architecture for gesture recognition in human-robot interaction.
• Developed an embedded AI interaction stack for a humanoid robot, combining real-time gesture recognition, computer vision and a fine-tuned LLM for domain-specific dialogue.
• Optimized on-device inference latency through GPU pipeline tuning, enabling real-time responses under hardware and runtime constraints.
• Implemented a Human-in-the-Loop framework to manage edge cases, improve robustness and support continuous model improvement.

Banque de France@banquedefranceData Scientist Intern

• Automated scraping workflows aggregating unstructured public and financial data from multiple sources.
• Combined OCR, Speech-to-Text and LLMs to digitize raw audio/visual content, feeding a downstream RAG system.

Comat Specific@groupe-comatMachine Learning Engineer

• Built a deep-learning OCR pipeline converting legacy hand-drawn 2D engineering sketches into structured machine-readable data, modernizing industrial workflows.

Embedded AI system combining real-time gesture recognition, computer vision and a fine-tuned LLM for domain-specific dialogue, with GPU latency tuning and Human-in-the-Loop robustness.

Lead author of an accepted HRI paper on an optimized real-time computer-vision architecture for gesture recognition, integrating MediaPipe landmarks, lightweight ML classifiers and low-latency robot actuation.

Deep-learning OCR pipeline converting legacy hand-drawn 2D engineering sketches into structured machine-readable data for industrial workflows.

Won a Kaggle-style machine learning challenge by building a classifier to assess whether SME loan applications should be accepted or denied, optimized with Macro F1-Score.

Analyzed a noisy three-variable dataset and selected a meaningful partitioning through exploratory analysis, cluster-number estimation and DBSCAN hyperparameter search.

Grade A - Top 10%. Optimization algorithms, Multicriteria Optimization, ML, Image Processing, Reinforcement Learning, Explainable AI, Probabilistic Modeling.

Machine Learning, cloud computing, large-scale data architectures.

Machine Learning, Deep Learning, Software Development, Embedded Systems.