University Internship
Pubblicato il 01-08-2026 - S&A Sistemi & Automazione in Italia
Senior AI Engineer
As a Senior AI Engineer , you will be responsible for developing and implementing Retrieval-Augmented Generation (RAG) systems and advanced feature extraction techniques to analyze textual, audio, and video content. You will collaborate with teams of data scientists, NLP engineers, and media analysts to create solutions that enhance information understanding and generation.
Work place: Rome
We offer: Permanent employment contract with a gross annual salary ranging from €38,000 to €42,000.
Responsibilities
- Design and develop AI agent pipelines for RAG, combining information retrieval and advanced generation with LLMs.
- Implement feature extraction techniques from images, video, and audio using CNNs, Transformers, and audio embedding models.
- Create pipelines for the ingestion, indexing, and retrieval of multimedia content, leveraging vector databases and advanced technologies such as FAISS, Pinecone, or Weaviate.
- Develop and optimize embedding models for multimedia and textual content, ensuring scalability and robustness of the solutions.
- Enhance the reasoning and contextualization capabilities of LLM models through fine-tuning, prompt engineering, and knowledge distillation.
- Optimize the integration of retrieval systems based on semantic search to respond to queries more accurately and efficiently.
What we offer
- A highly collaborative team environment with a strong focus on technical quality.
- 2 days per week of remote work.
- Tailored growth paths and ongoing training.
- Opportunity to work directly on-site with clients on structured projects.
Requirements
- 8+ years of working experience.
- 3+ years of experience developing AI solutions with a focus on NLP, multimodal AI, or feature extraction.
- Excellent knowledge of Python and major AI libraries, including Hugging Face, LangChain, PyTorch, and TensorFlow.
- Experience with RAG architectures, vector databases, and semantic search techniques.
- Familiarity with preprocessing, tok