Hello, I'm Malvin
Data Scientist & AI Engineer
Building cutting-edge LLM and RAG solutions that empower engineering teams to automate and optimize complex workflows. Let's bring AI innovation to life together!
Building cutting-edge LLM and RAG solutions that empower engineering teams to automate and optimize complex workflows. Let's bring AI innovation to life together!
I believe AI should augment human potential. My work focuses on creating robust, production-ready AI workflows—especially in Generative AI and Retrieval-Augmented Generation—to help teams solve real-world challenges faster.
Currently exploring content automation using GenAI pipelines for video creation (e.g., auto-generating TikTok/YouTube Shorts with AI voice, script, and video search from structured inputs).
Politecnico di Milano
Exchange Semester in AI (2024 - 2025)
🏆 Finished 8th in the Polimi Recommender System Challenge (2024)
ESILV (Léonard de Vinci Engineering School)
MSc Data Science & AI (2022 - 2025)
Safran - AI Intern / Generative AI & RAG
Feb 2025 - July 2025
Developed a RAG system for engine impact analyses, used hybrid retrieval (FAISS + BM25) & LLMs (Mistral, LLaMA).
Thales - R&D Intern in Applied Math & AI
Apr 2024 - Aug 2024
Evaluated AI models for fatigue detection in crew monitoring, improved detection speed & reliability.
A Retrieval-Augmented Generation system for analyzing CFM56 & LEAP engine incidents. Integrates FAISS + BM25 to retrieve past case data for faster troubleshooting.
R&D project for Thales to detect crew fatigue via face and motion analysis. Implemented model hyperparameter tuning and Docker-based workflows.
Exploring embeddings-based search and few-shot prompt engineering with LLMs like Mistral and LLaMA to create domain-specific chatbots.
GenAI-powered app that automates creation of TikTok & YouTube Shorts from structured prompts (CSV). The pipeline generates voice-over, script, and fetches relevant footage online to assemble full videos autonomously.