AI Engineer

Scania is currently undergoing an exciting transformation from a traditional truck manufacturer to a provider of complete sustainable transport solutions. IT is a crucial part of this transformation as Scania’s success depends on flexible and efficient IT solutions that support current and future business requirements.  Our vision   To drive the shift by democratizing the access to knowledge and artificial intelligence fostering the transition towards a sustainable transport system. We are currently working with key initiatives ongoing to leverage AI not only for productivity gains but to fundamentally enhance and transform our current processes or even create new business opportunities across Scania and TRATON. Your role • Lead and contribute to the technical development of the Generative AI platform. • Contribute towards the strategy and roadmap for the Generative AI area, together with the team. • Building, testing and deploying AI models from scratch or using the existing platform. • Conducting statistical analysis and interpreting the results to optimize the performance of AI models and systems • Developing self-service Generative AI services • Support the implementation of LLMOps within the team. • Support and contribute to setting AI best practices and methodologies. • Work closely with cross-functional teams to capture the business needs and develop valuable solutions. • Serve as ambassador for the area of generative AI and the AI platform both internally and externally, sharing knowledge and answering questions. Must have skills and qualifications: • Understanding of Transformer Architectures (BERT, GPT, T5 etc), Named Entity Recognition (NER) using Spacy Models and libraries like NLTK,Langchain Loaders,Sentiment Analysis, Topic Modeling, Text Summarization • Experience in creating optimized RAG applications:  o Dense Retrieval mechanisms (e.g., DPR, ColBERT), o Sparse Retrieval (BM25, TF-IDF etc.),  o Hybrid Retrieval (HyDE), o Embedding Techniques and retrievers, o Vector Databases (Faiss, Pinecone, others), o Query Understanding and Expansion (Chain of Thoughts, Tree-of-Thoughts, etc.), o Context-Aware Retrieval, o Evaluation Metrics (Relevance, Factual Accuracy, Coherence, Diversity, Perplexity etc), o Prompts (e.g., QuestionAnswerPrompt, RefinePrompt), o Optimizers (e.g., OptimizerChain)  • Experience with fine tuning mechanisms • Familiarity with the latest models such as GPT-3.5/4/4o, LLAMA, and Domain-Specific Fine-Tuning. • Experience in building self-supervised corpora, Q&A pair templates, RHLF, prompt engineering templates. • Data Collection and Curation, Preprocessing, Exploratory Data Analysis (EDA), Feature Engineering, Data Augmentation and noise removal Techniques, familiarity with Large-Scale Datasets, Data Lakes • Model Versioning, Experiment Tracking (Mlflow/Kubeflow, Weights & Biases), CI/CD for ML, Monitoring and Logging, Reproducibility and Provenance, Containerization (Docker), Orchestration (Kubernetes) • Familiarity with Multi-Agent AI frameworks and agentic patterns Programming & Tools:  • Python, • PyTorch/TensorFlow, • Keras, • Hugging Face, • Transformers, • Git, • Docker, • SQL Good to have: • Knowledge about Cloud Platforms (AWS, Azure), Model Serving (Flask, FastAPI), Model Compression • Knowledge and familiarity with Traditional machine learning