machine learning engineer for large-scale RAG platforms

EPAM develops enterprise software products, open source solutions, and accelerators.

• Deploy and manage Milvus vector databases, including schema design and index tuning with HNSW and IVF-FLAT;
• Build embedding and LLM framework pipelines using OpenAI API, Hugging Face, or Cohere;
• Manage Kubernetes clusters, Helm charts, and containerized microservices for scalable orchestration;
• Implement Docker containerization with multi-stage builds and registry management;
• Develop production-level applications in Python along with Go, Java, or C++;
• Integrate object storage systems including AWS S3, MinIO, or Google Cloud Storage;
• Support large-scale RAG applications and multi-agent platforms;
• Optimize compute and inference through GPU scheduling, resource optimization, and inference acceleration;
• Drive search optimization with hybrid search, metadata filtering, and index tuning;
• Collaborate with the team to deliver high-quality solutions.

• B.Tech/B.E in Engineering and 5+ years of relevant experience;
• Expertise in Milvus deployment, schema design, and index tuning with HNSW and IVF-FLAT;
• Familiarity with Qdrant, Pinecone, Weaviate, PGVector, or Chroma;
• Proficiency in OpenAI API, Hugging Face, or Cohere for embeddings and LLMs;
• Skills in Kubernetes cluster management, Helm charts, and containerized microservices;
• Competency in Docker containerization, multi-stage builds, and registry management;
• Production-level proficiency in Python along with Go, Java, or C++;
• Knowledge of object storage integration including AWS S3, MinIO, or Google Cloud Storage;
• Excellent verbal and written communication skills;
• Nice to have: background in supporting large-scale RAG applications and multi-agent platforms, familiarity with LangChain, LlamaIndex, or custom LLM orchestration pipelines, understanding of AI observability through LLM evaluation, governance, tracing, and monitoring tools, knowledge of CI/CD pipelines, Infrastructure-as-Code, and cloud-native deployment practices, prior work experience in the Oil and Gas industry along with Dataiku DSS and SRE practices.

• No conditions specified