Empresa:

ILI.DIGITAL GmbH

Descrição da Função

About the Role

We are hiring a Generative AI Engineer who brings together deep knowledge of natural language processing (NLP), cloud services, and document processing. You'll design and implement scalable GenAI pipelines for extracting intelligence from PDFs, Word documents, and unstructured data sources. The role requires hands-on experience with both modern LLMs and classical NLP algorithms, along with an understanding of traditional deep learning architectures like RNNs and LSTMs, particularly in TensorFlow or PyTorch environments. You'll work across model development, cloud integration, and API delivery, turning unstructured content into structured insight using the latest Generative AI + Document AI technologies.

Key Responsibilities

• Develop and fine-tune LLM-based systems for text generation, summarization, Q&A, and document intelligence.
• Work with PDF, DOCX, and scanned documents using parsing and OCR libraries.
• Apply classical NLP algorithms (e.g., TF-IDF, Named Entity Recognition, POS tagging, dependency parsing).
• Implement and optimize deep learning models including RNNs and LSTMs, especially for sequence modeling and text classification tasks.
• Build retrieval-augmented generation (RAG) pipelines using vector databases and embedding models.
• Use OCR tools like Azure Document Intelligence (Form Recognizer), Amazon Textract, Google Document AI, or open-source alternatives.
• Deploy GenAI services on cloud platforms (Azure, AWS, GCP).
• Collaborate across teams to productionize models with API endpoints, logging, and monitoring.

Must-Have Qualifications

• Bachelor's or Master's in Computer Science, AI, or related field.
• 3+ years experience in NLP/ML, including classical and deep learning methods.
• Strong experience with TensorFlow and/or PyTorch, including building and training RNN/LSTM architectures.
• Solid grasp of classical NLP algorithms: TF-IDF, BM25, NER, POS Tagging, Text Classification, Topic Modeling (LDA), Dependency Parsing, Lemmatization, Stemming.
• Hands-on experience with PDF and Word parsing (PyMuPDF, pdfminer, python-docx).
• Experience with OCR modules (Azure Document Intelligence, Amazon Textract, Google Document AI, Open Source Alternative).
• Proficient in Python (Hugging Face Transformers, spaCy, scikit-learn).
• Cloud experience with AWS, Azure, or GCP.
• Familiar with MLOps, FastAPI, and Docker.

Nice to Have

• Experience with LangChain, LlamaIndex, or Haystack.
• Understanding of multi-modal models (text + layout + images).
• Familiarity with RLHF, hallucination prevention, or LLM security.
• Exposure to Kubernetes, CI/CD, and Terraform.

Observações

Lisboa (Portugal)