Job Summary
"Role Overview:
Design, implement, and optimize Retrieval-Augmented Generation (RAG) systems for enterprise AI applications. Enable LLMs to access and utilize internal knowledge bases, ensuring accurate, context-rich, and verifiable outputs.
Key Responsibilities
Key Responsibilities:
Architect and build RAG pipelines integrating LLMs with enterprise data sources (documents, databases, APIs).
Develop and maintain retrieval components (vector databases, embedding models, search algorithms).
Implement data ingestion, indexing, and segmentation for large-scale, multi-modal content (PDFs, DOCX, PPTX, images, etc.).
Ensure robust security, compliance, and access controls for sensitive enterprise data.
Optimize RAG system performance, reliability, and scalability for production use.
Collaborate with data engineers, ML engineers, and business stakeholders to deliver end-to-end RAG solutions.
Evaluate and improve RAG system accuracy, relevance, and user experience.
Document processes, architectures, and best practices for knowledge sharing.
Skill Requirements
Required Skills:
5+ years in data engineering, ML engineering, or AI application development.
Strong proficiency in Python, SQL, and data processing frameworks.
Hands-on experience with vector databases (Pinecone, Weaviate, FAISS, OpenSearch), graph databases (Neo4j, CosmosDB), and RAG system implementation.
Familiarity with LLMs (GPT-4, Claude, Gemini, Llama), prompt engineering, and information retrieval.
Experience with enterprise content management systems and ETL/ELT pipeline development.
Knowledge of security, compliance (e.g., HIPAA), and governance for enterprise data.
Excellent problem-solving, communication, and documentation skills.
Preferred:
Experience with large-scale embedding, retrieval, and ranking systems.
Exposure to GenAI frameworks (LangChain, Hugging Face, Vertex AI, etc.).
Background in healthcare, finance, or regulated industries is a plus.
Sample Use Cases:
Building enterprise RAG systems for customer support, knowledge management, and decision support.
Architecting robust, scalable RAG pipelines for multi-modal and multi-source data.
Implementing secure, compliant RAG solutions for regulated environments."