Data Engineer II

Description

eSimplicity is a modern digital services company that partners with government agencies to improve the lives and protect the well-being of all Americans, from veterans and service members to children, families, and seniors. Our engineers, designers, and strategists cut through complexity to create intuitive products and services that equip federal agencies with solutions to courageously transform today for a better tomorrow.

The LLM Specialist will drive the design, development, and operationalization of advanced large‑language‑model capabilities across a cloud‑based analytics ecosystem. This role leads innovation efforts around cutting‑edge AI, owning the architecture and strategy for fine‑tuning, retrieval‑augmented generation (RAG), agentic frameworks, and domain‑specific model adaptation. The specialist will guide the development of high‑impact prototypes, oversee the evolution of scalable LLM pipelines, and ensure robust governance, security, and performance across all model implementations.

Partnering with engineering, product, and data teams, this position provides technical leadership, evaluates emerging LLM technologies, sets best practices, and helps drive transformation through the practical, safe, and effective deployment of generative AI.

• All candidates must pass public trust clearance through the U.S. Federal Government. This requires candidates to either be U.S. citizens or pass clearance through the Foreign National Government System which will require that candidates have lived within the United States for at least 3 out of the previous 5 years, have a valid and non‑expired passport from their country of birth and appropriate VISA/work permit documentation.
• Bachelor’s Degree and 5+ years of previous systems engineering experience
• Experience developing and working with large language models (LLMs), transformer‑based architectures, and generative AI solutions.
• Experience fine‑tuning LLMs, applying parameter‑efficient training methods (e.g., LoRA, PEFT), and developing effective prompt engineering strategies.
• Experience designing, implementing, and optimizing Retrieval‑Augmented Generation (RAG) solutions, including embeddings, retrieval workflows, vector databases, and search optimization.
• Hands‑on experience with LLM development frameworks and orchestration tools such as Lang Chain, Llama Index, or similar technologies.
• Strong Python programming skills with experience building, testing, and deploying AI/ML applications.
• Experience working with distributed computing environments, GPU‑accelerated workloads, or large‑scale model training and inference.
• Experience designing, deploying, and supporting AI/ML solutions in cloud environments such as Microsoft Azure, Amazon Web Services (AWS), Google Cloud Platform (GCP), or similar platforms.
• Knowledge of MLOps and LLMOps practices, including source control, CI/CD pipelines, automated testing, monitoring, performance optimization, and model governance.
• Ability to lead technical discussions, collaborate effectively with cross‑functional teams, mentor team members, and communicate complex technical concepts to both technical and non‑technical audiences.

• Experience implementing multi‑agent or agentic AI systems for task automation and reasoning.
• Familiarity with LLM evaluation frameworks, structured benchmarking, or human‑in‑the‑loop refinement methods (e.g., RLHF‑style workflows).
• Expertise with advanced retrieval techniques such as hybrid search, graph retrieval, or long‑context optimization.
• Experience optimizing model inference through quantization, model compression, or model distillation.
• Background integrating LLM services with large‑scale analytics environments (e.g., Databricks, Snowflake, Spark).
• Strong skills in exploratory data analysis, feature engineering, and data modeling to support domain‑specific LLM customization.
• Experience developing innovative prototypes or POCs that leverage state‑of‑the‑art generative AI approaches.
• Exposure to emerging architectures such as mixture‑of‑experts models, long‑context transformers, or experimental generative frameworks.