Remote AI Engineer – LLM/RAG for Government

GDH; North Carolina; Remoto; USD 110.000 - 116.000; Tempo pieno

Descrizione del lavoro

GDH is seeking a contract AI Engineer to develop RAG and document-processing pipelines in a federal context. The role focuses on building scalable, secure AI on a greenfield platform and delivering a near-term product demonstration.

You will implement privacy-preserving synthetic data, manage deployments of open-weight LLMs, and collaborate across teams to meet security requirements. Remote work within the United States is allowed.

Competenze

• 8+ years in ML/AI or data systems with deliverable LLM/RAG projects beyond notebooks.
• Hands-on Databricks development, deployment and management.
• Processing data at scale including OCR and parsing of mixed-quality sources.
• Privacy-preserving synthetic data from sensitive sources; awareness of re-identification risks.
• Proficiency in Python and related data science tools.
• Experience in government or defense contracting environments.

Mansioni

• Develop retrieval-augmented generation (RAG) and document-processing pipelines using Databricks lakehouse, OCR, chunking, embedding and retrieval methods.
• Build LLM workflows for summarization, structured data extraction, and evidence-grounded text generation with source attribution.
• Generate synthetic document corpora that meet fidelity and quality variation criteria.
• Establish evaluation frameworks for retrieval quality, grounding, hallucination rates and structured output validity.
• Package deliverables as jobs and asset bundles; track progress with MLflow; maintain comprehensive documentation.
• Implement privacy-preserving synthetic data generation techniques for sensitive sources and re-identification risk awareness.
• Ensure secure local deployment of open-weight LLMs (e.g., TGI, Ollama, llama.cpp) in isolated or air-gapped environments.
• Collaborate with cross-functional teams to integrate AI solutions within federal security and compliance standards.
• Monitor and report model performance metrics with clear insights for evaluation and improvement.

Conoscenze

Python

Databricks

RAG systems

LLM workflows

OCR processing

Data extraction

Security/compliance

Strumenti

vLLM

TGI

Ollama

llama.cpp

Descrizione del lavoro

GDH is seeking a contract AI Engineer to develop RAG and document-processing pipelines in a federal context. The role focuses on building scalable, secure AI on a greenfield platform and delivering a near-term product demonstration.

You will implement privacy-preserving synthetic data, manage deployments of open-weight LLMs, and collaborate across teams to meet security requirements. Remote work within the United States is allowed.