Description

Design and implement AI systems focusing on generative retrieval and agentic workflows.

Responsibilities

• Build RAG pipelines on Google Cloud/Vertex AI, managing chunking, embeddings, indexing, retrieval, reranking, and grounding.
• Develop agentic workflows using Python-first frameworks, implementing tool use, planning, reflection, and structured outputs.
• Create ETL processes from diverse sources (PDFs, docs, webpages) while defining metadata strategy and access control.
• Define and execute system evaluations to measure retrieval metrics, answer quality, and hallucination rates.
• Ship production-ready systems, managing APIs, observability, cost optimization, CI/CD, and security.

Required Skills

• 5+ years developing complex AI/ML applications.
• Strong Python proficiency with clean architecture, async patterns, typing, and testing.
• Proven experience building RAG solutions, including hybrid search, reranking, and prompt design.
• Hands-on experience with Vertex AI and GCP fundamentals (IAM, logging, Cloud Run/GKE).
• Experience with agentic frameworks (LangGraph/LangChain, LlamaIndex, AutoGen) and tool calling.
• Solid knowledge of vector search and deployment with vector DBs (e.g., Pinecone, pgvector).
• Familiarity with graph data modeling and querying languages (Cypher/Gremlin/SPARQL).
• Experience with ETL for unstructured content and secure-by-design CI/CD practices.

Preferred Skills

• Experience with knowledge graphs for RAG, including entity linking and retrieval fusion.
• Familiarity with streaming/messaging systems (Pub/Sub, Kafka) or Document AI.

Key Skills

Education

Any Gradute