AI Engineer at NielsenIQ

AI Engineer (Mid‑Level) We’re looking for a Mid-Level AI Engineer who is passionate about building secure, scalable, and production‑ready AI systems. In this role, you’ll design, deploy, and optimize machine learning and generative AI solutions that power analytics, automation, and intelligent products. You’ll work across Python, cloud platforms, vector databases, LLM frameworks, and modern MLOps practices — treating testing, reliability, and security as foundational pillars.  Key Responsibilities AI/ML Model Development  • Build and optimize machine learning and LLM-based models (classification, NLP, embeddings, generative AI).  • Develop reusable model pipelines in Python using libraries like scikit‑learn, PyTorch, TensorFlow, or HuggingFace.  • Implement feature engineering, vectorization, and model evaluation frameworks. Generative AI & LLM Engineering  • Fine‑tune, evaluate, and deploy LLMs using frameworks such as HuggingFace Transformers, LangChain, LlamaIndex, or cloud-native AI services.  • Build RAG (Retrieval-Augmented Generation) pipelines using vector databases (Pinecone/Weaviate/FAISS/Snowflake Cortex Search).  • Implement prompt engineering, prompt orchestration, and prompt quality testing. MLOps & CI/CD  • Build automated training, deployment, and monitoring pipelines using CI/CD (GitHub Actions/Azure DevOps/GitLab CI).  • Manage model versioning, lineage, and artifact tracking (MLflow, Weights & Biases).  • Implement reproducibility, environment management, and dependency controls. Cloud & Deployment  • Deploy AI services to AWS/Azure/GCP using serverless compute or containerized microservices (Lambda, Azure Functions, Cloud Run, Kubernetes).  • Optimize compute workloads, costs, and auto-scaling strategies.  • Integrate models into production systems via APIs, SDKs, or event-driven flows. Security, Reliability & Observability  • Apply DevSecOps practices: secure model endpoints, manage secrets, enforce least‑privilege IAM.  • Implement monitoring for drift, model decay, model failures, and performance anomalies.  • Contribute to logging, metrics, dashboards, and alerting for AI workloads. Collaboration & Documentation  • Partner with Data Engineering, Product, and Analytics teams to integrate models into applications and data pipelines.  • Write clear documentation (model cards, evaluation reports, design documents, runbooks).