Lead Data Scientist & Full-Stack AI Engineer based in Berlin, Germany 🇩🇪

I build end-to-end AI systems that solve real-world problems—from agentic LLM workflows to predictive maintenance in manufacturing, healthcare optimization, and industrial digital twins.

• Agentic AI Systems: Designing and deploying LLM-powered agents with complex reasoning, multi-step workflows, and real-time decision-making (DSPy, LangGraph, LangChain, LlamaIndex, Azure AI, OpenAI)
• Predictive Analytics: RUL estimation, anomaly detection, and time-series forecasting for manufacturing and aerospace
• Optimization & Operations Research: Meta-heuristics, gradient-free optimization, warehouse management, and production scheduling
• Computer Vision: Industrial defect detection, quality control, and automated visual inspection
• Digital Twins & Simulation: Building virtual replicas for scenario analysis, optimization, and reinforcement learning-driven control
• MLOps & Production ML: Full-stack ML deployment, monitoring, A/B testing, and scalable inference pipelines

Languages: Python, SQL, JavaScript

ML/DL Frameworks: PyTorch, TensorFlow, scikit-learn, XGBoost, LightGBM

AI & LLMs: DSPy, LangGraph, LangChain, LlamaIndex, OpenAI API, Azure AI, Hugging Face Transformers

MLOps: MLflow, DVC, Weights & Biases, Kubeflow, Docker, Kubernetes

Backend & APIs: FastAPI, Flask, Django, Node.js

Data & Cloud: PostgreSQL, Redis, Azure, AWS, GCP, Apache Spark

Optimization: Genetic Algorithms, Simulated Annealing, Particle Swarm, Mixed-Integer Programming

Current & Recent Roles:

• Lead Data Scientist at indx.com — Building agentic AI systems for enterprise automation
• AI/ML Engineer at Jampp — Predictive models for mobile advertising optimization
• Research & Development in manufacturing optimization, healthcare analytics, and aerospace predictive maintenance

Academic Background:

• Master's Degree in Data Science — Università degli Studi di Padova 🇮🇹
• Master's Degree in Computer Science — Universidad Nacional de La Plata 🇦🇷

Automated SAP ticket resolution system with LLM-powered root-cause analysis, stakeholder communication, and monitoring UI.
Stack: LangGraph · Azure AI · FastAPI

Stock management and power core order optimization for slitted coils, minimizing scrap while meeting production constraints.
Stack: Python · Meta-heuristics

Deep learning models for Remaining Useful Life prediction to improve maintenance planning and reduce downtime.
Stack: PyTorch · MLflow

High-precision anomaly detection on component images for automated quality control in manufacturing.
Stack: CNNs · Computer Vision

Gradient-free optimization for product planning and scheduling across industrial digital twins.
Stack: Python · Simulation

Reinforcement learning controller trained on synthetic data from digital twins to optimize energy efficiency.
Stack: RL · Simulation

I share ML experiments, tutorials, and project deep-dives on my YouTube channel. Topics include:

• Agentic AI workflows and LLM engineering
• Predictive maintenance and time-series forecasting
• Computer vision and deep learning techniques
• Optimization algorithms and meta-heuristics

I actively contribute to the research community with publications in:

• Predictive Maintenance & Reliability Engineering
• Machine Learning for Industrial Systems
• Optimization & Operations Research
• Computer Vision & Anomaly Detection

Check out my full publication list on Google Scholar.

• Spanish — Native
• English — Fluent
• Italian — Fluent
• German — B1

I'm always open to collaborating on interesting projects, discussing research ideas, or exploring opportunities in AI/ML engineering.

• LinkedIn: luciano-lorenti
• Email: [email protected]
• Google Scholar: View Publications
• YouTube: @lolo.lorenti

⭐️ Feel free to explore my repositories and don't hesitate to reach out!