Applied AI Engineer
Core
Build and maintain the retrieval, generation, and evaluation systems for Arthur, a software product that analyzes public records, regulatory filings, and community data to provide clients with traceable analytics and insights.
Role type
Senior Applied AI Engineer (LLM systems)
Builds
Retrieval layer, generation pipeline for research analyses, classifiers, and entity extraction systems for Arthur.
Domain
Public policy, infrastructure development, and civic data analytics
Deliverable
production ML models | product features
Required skills
Python, PostgreSQL, LLM prompt design, structured output schemas, evaluation of generated text, retrieval-augmented generation (RAG), model impact measurement
Preferred skills
pgvector, LoRA fine-tuning, vLLM, NVIDIA hardware, classification systems, entity-linking, public records processing
Technologies
Python, PostgreSQL, vLLM, pgvector, LLMs
Responsibilities
Build and maintain the retrieval layer including chunking, embeddings, indexing, reranking, and citations. Develop the generation pipeline for research analyses with structured outputs. Create classifiers for signal type, impact, and stakeholder stance. Develop entity extraction and linking systems. Maintain and run evaluation sets against prompt and model changes. Prepare analyst-approved briefs as training data and run fine-tuning jobs. Collaborate with analysts to translate corrections into system improvements. Document prompts, schemas, and operating procedures.
Seniority
Senior, hands-on IC