Senior RAG AI Engineer — Source-Grounded Legal Tech

LexisNexis Legal; Ciudad de México; On-site; MXN 900,000 - 1,200,000; Full time

Job summary

LexisNexis Legal & Professional is seeking an AI Engineer focused on Retrieval-Augmented Generation (RAG) to design, build, and optimize AI-powered experiences. You will handle document ingestion, embedding generation, and source-grounded responses within a rigorous evaluation framework.

You will work with cutting-edge vector search tech to ensure accurate citations and low hallucinations, contributing to trusted legal and professional information products.

Qualifications

• Hands-on experience with Retrieval-Augmented Generation (RAG) applications.
• Strong understanding of Generative AI and LLM capabilities.
• Experience with vector search technologies and eval datasets.

Responsibilities

• Design, build, and optimize Retrieval-Augmented Generation (RAG) applications.
• Develop document ingestion, chunking, embedding generation, retrieval, reranking, and source-grounded responses.
• Design and operate vector search solutions using modern vector database technologies.
• Validate retrieval quality, citation accuracy, relevance, coverage, and factual grounding using reproducible evaluation datasets.
• Collaborate with cross-functional teams to improve response quality and reduce hallucinations.
• Implement hybrid search strategies combining keyword, metadata, semantic/vector search, and reranking.
• Monitor and evaluate retrieval performance, latency, cost, and AI system quality.
• Support the development of scalable and responsible Generative AI solutions across legal and professional information products.

Skills

RAG apps

Generative AI

LLM evaluation

Vector search

Reproducible evals

Tools

Qdrant

Pinecone

pgvector

LanceDB

OpenSearch vector

Job description

LexisNexis Legal & Professional is seeking an AI Engineer focused on Retrieval-Augmented Generation (RAG) to design, build, and optimize AI-powered experiences. You will handle document ingestion, embedding generation, and source-grounded responses within a rigorous evaluation framework.

You will work with cutting-edge vector search tech to ensure accurate citations and low hallucinations, contributing to trusted legal and professional information products.