Alejandro Rosas@alejandro_rosas
0
Events attended
0
Submissions made
1 year of experience
About me
Applied mathematician (PhD, Institut Polytechnique de Paris) transitioning into AI engineering and data science from a background in high-performance scientific computing. Professional experience at INRIA developing and validating 3D numerical simulation pipelines in Python and Matlab. Self-directed specialisation through an intensive DataCamp programme (170+ courses, 10 career certifications), covering LLM application development, RAG pipelines, agentic systems with LangChain and LangGraph, fine-tuning with Llama 3, statistical modelling, predictive analytics, time series analysis, machine learning, and end-to-end MLOps pipelines — validated across production-grade projects spanning both disciplines. Quaestor, a cited-answer RAG system for SEC filings, achieved a 300% answer rate improvement with confidence-gated hallucination detection and a full RAGAS/Langfuse observability stack. A Regime-Aware Volatility Forecasting system (GARCH + HMM + XGBoost) delivered a statistically significant 10.28% RMSE improvement with full DVC/MLflow traceability.
