Are you interested in building software that creates meaningful social impact? Every solution this company develops is deployed directly to production, delivering tangible, real-world benefits within the healthcare sector.This Irish software company develops AI-driven services using large language models, combining in-house components (30%) with third-party AI APIs and platforms (70%), including Cursor integrated with Claude models. While AI can be simple and intuitive on its own, this role involves working with a sophisticated AI system that introduces technical complexity and encourages innovative problem-solving. If you are a team player who is passionate about technology, an effective communicator, and enjoys exploring creative ways to solve problems while tackling complex challenges, this is an environment where you can thrive and make a meaningful impact.Essential Skill-set Machine Learning & Frameworks PyTorch (3+ years)Essential: Strong experience in model development and training. Build custom models, implement data pipelines, and fine-tune pretrained models.Nice-to-have: Mixed precision, quantization, distributed or large-scale training.Transformer-based NLP Models (2+ years)•Hands-on experience with tasks such as understanding, classifying, extracting, summarizing, or embedding text.•Practical experience using Hugging Face Transformers library; familiarity with BERT is advantageous.Production Deployment (2+ years): Experience deploying Transformer/NLP models (e.g., Hugging Face, BERT) in a production environment.MLOps & Deployment•Solid understanding of MLOps principles: model versioning, experiment tracking, and CI/CD for ML.•Essential: versioning and tracking models, automated CI/CD pipelines (e.g., GitHub Actions), deployment and scaling using Docker.•Third-party LLM Integration (1+ year): Experience integrating LLMs via APIs in production environments.Backend & Cloud Expertise•Proficient in Python for backend services and ML pipelines; experience with frameworks like FastAPI, Flask, or similar.•Hands-on commercial experience with cloud ML platforms; They use AWS Bedrock, but Google Vertex AI or Azure OpenAI experience is acceptable.What will you be doing in this roleNLP API Development & Maintenance (Approx 40%) Delivering new features (80%) and handling ongoing maintenance and improvements (20%)MLOps (Approx 20%): Designing and maintaining data pipelines, automating deployment workflows, and managing model versioningASR Model Fine-Tuning (Approx 15%) Fine-tuning (ASR) models to improve accuracy and performanceLLM Fine-Tuning for ASR (Approx 15%) Adapting and fine-tuning large language models for ASR-related tasksR&D and Experimentation (Approx 20%) Research, experimentation, and prototyping for fine-tuning LLMs and training NLP modelsLocation: Dublin City Centre, hybrid