The Group Advanced Analytics and Automation Solutions Unit is in charge of driving the execution of key Artificial Intelligence projects at Group level, related both to AI-driven automation and AI recommendations/predictions and support local Business Units to implement local projects. The unit is also responsible for designing, implementing, and continuously improving a library of central shared analytics / automation services, leveraging cutting-edge technologies, algorithms and methodologies, available for the BUs through a central platform.
In our team, the Data Scientist is in charge of identifying, designing and developing artificial intelligence solutions through text analytics, image analytics and data analytics technologies, with the aim to automate end-to-end discretionary processes, such as claim settlement, underwriting or document processing, or to provide recommendations within human-based processes, such as fraud detection or churn prediction.
As a Data Scientist, you will take part in international AI projects and initiatives in collaboration with our +50 Business Units, having the opportunity to gain full ownership on AI applications, implementation and delivery in the different sectors and countries where Generali operates.
On top of the project delivery, you will have the chance to experiment the most innovative technologies and approaches in order to continuously enrich the Group library of analytics / automation services.
This includes for example:
-
Gradient Boosting, Neural Networks, Unsupervised learning and Optimization techniques for core insurance use cases like pricing and fraud detection
-
Transformers for Token Classification, Question Answering, Semantic Search, Relation Extraction and Dependency parsing in different languages; Multi-task learning and Fine-tuning on internal data assets; Prompt Engineering with LLMs
-
State-of-the-Art algorithms, frameworks and projects for Computer Vision, metric learning and time-series Analysis (for example Hugging Face, Detectron2, MMLab) and model optimization for deploy at scale (for example NVIDIA Triton, ONNX, TorchScript)
-
Synthetic data generation (including 3D computer graphics and Generative Adversarial Networks - GANs)
-
GPU and Multi-GPU trainings
Main tasks and responsibilities:
-
Support the project set-up, the analysis of the as-is processes, the creation of a business case, and the design of the target end-to-end solution, including the models and data needed to maximize the business impact.
-
Develop end-to-end AI projects, including machine learning models, optimization techniques and algorithms, from the data pre-processing, to the code-base creation and model training, evaluation and test.
-
Present and explain the projects results and the solution’s performances to the business stakeholders
-
Research bleeding-edge techniques and tools in machine learning/deep learning/artificial intelligence to bring new research topics and methodologies from paper to production, and contribute to the library of ready-to-use shared models / algorithms used by all Business Units
Our ideal candidate will meet the following requirements:
-
Strong problem-solving skills, creativity, and determination
-
Excellent analytical and deductive skills
-
Experience in Python and other programming languages (R, Java, C++, …) and Data Science libraries (Pandas, Scikit-learn, Numpy, Pytorch, LightGBM, Tensorflow)
-
Experience in designing, developing and evaluating Statistical modeling or Machine Learning models (clustering, classification, regression, forecasting, …)
-
(Nice-to-have) Experience in Cloud Development (preferable AWS), ML Ops, Containerization multi-GPU/multi-node training, model optimization (pruning, quantization, distillation, …) and the most famous Deep Learning frameworks (Tensorflow and Pytorch) or libraries (Hugging Face, Detectron2, MMLab, …)
-
Ability to work in a team, operational autonomy, proactivity, high level of commitment and result-orientation
-
Master’s degree in scientific disciplines (Mathematics, Computer Science, Engineering, Economics, Physics)
-
Good knowledge of written and spoken English (B2);
-
Basic knowledge of IT architecture (nice-to-have)