AI Engineer
roche
Job Description
- 5–8 years of experience in ML/NLP, applied AI, data engineering, or a related technical field, including at least 2–3 years of hands-on experience developing Generative AI or LLM-based solutions.
- Practical experience with NLP, transformer-based models, and LLM application development using PyTorch, TensorFlow, or Hugging Face Transformers.
- Hands-on experience with Generative AI application and agent-orchestration frameworks such as LangChain, LangGraph, LlamaIndex, or CrewAI.
- Experience with vector databases and vector search technologies such as Pinecone, Chroma, or PostgreSQL with pgvector.
- Strong proficiency in Python and SQL, with experience preparing and transforming data for model development, evaluation, and production AI workflows.
- Experience building retrieval pipelines involving document chunking, embedding generation, vector indexing, metadata filtering, similarity search, and result reranking.
- Strong prompt and context engineering skills, including structured-output design, tool calling, and grounding responses in approved data sources.
- Experience implementing input and output guardrails, fallback mechanisms, automated evaluations, and regression tests for LLM applications.
- Expertise in deploying and operating production AI/ML and Generative AI solutions on AWS or another major cloud platform, including monitoring, scaling, security, reliability, and cost optimization.
- Proficiency in Git, Docker, automated CI/CD pipelines, unit and integration testing, REST API development, and microservices architecture.
- Experience with LLM observability platforms such as Langfuse or LangSmith to trace and monitor model calls, agent workflows, tool usage, retrieval steps, errors, latency, and cost.
- Experience working in SAFe/scaled Agile environments.
- Demonstrated ability to collaborate across product, engineering, data, cybersecurity, privacy, quality, regulatory, and domain teams while identifying risks and dependencies early.
- Strong analytical, written, and verbal communication skills, with the ability to explain technical trade-offs clearly to technical and non-technical stakeholders.
- Experience applying security-by-design and privacy-by-design principles to AI systems handling sensitive data, including least-privilege access controls, encryption in transit and at rest, secrets management, audit logging, data minimization, and secure access to model endpoints.