Lead LLM / Generative AI Engineer

Job Description

Job DescriptionSalary:

PM Solutions is standing up an internal Large Language Model (LLM) environment to showcase and operationalize our AI solutions for federal customers. Were looking for ahands-on Lead LLM / Generative AI Engineerto design, build, and operate a secure internal LLM platform that powers demos, prototypes, and early-stage products for our government clients. This person will be the technical lead responsible for everything from model selection and architecture design to deployment, evaluation, and demo applications that our BD and delivery teams can put in front of customers. This is a hybrid position that requires the candidate to be on-site in the DC metro area at least two times per week.

Key Responsibilities

Platform & Architecture

• Design and build an internal LLM-based AI platform that runs in secure cloud environments (e.g., Azure Government, AWS GovCloud, or other FedRAMP-authorized environments).
• Evaluate and integrate commercial and open-source models (OpenAI / Azure OpenAI, Anthropic, Llama, etc.) based on performance, licensing, security, and cost.
• Implement Retrieval-Augmented Generation (RAG) pipelines to safely use internal documents and data in demos and prototypes.

Model Development & Evaluation

• Fine-tune or adapt LLMs for specific government use cases (document summarization, Q&A, workflow automation, knowledge assistants, etc.).
• Build robust evaluation frameworks and test suites to measure quality, latency, hallucinations, and safety of model outputs.
• Optimize models and inference pipelines for performance and cost (quantization, caching, batching, etc.).

Data, Integrations & Demos

• Design secure data ingestion pipelines and vector stores to support demos (e.g., pgvector, Pinecone, Weaviate, Redis, etc.).
• Build demo applications and proof-of-concept workflows (e.g., web apps, chatbots, copilots) that integrate with the internal LLM platform.
• Partner with business development, solutions, and proposal teams to rapidly prototype customer-specific demos for pursuits and orals.

Security, Compliance & Governance

• Work with leadership to ensure the AI environment aligns with federal security requirements (e.g., FedRAMP, NIST SP 800-171, DFARS 252.204-7012, HIPAA/PHI where applicable).
• Implement guardrails, content filters, access controls, and logging around model inputs/outputs to protect CUI, PII, and other sensitive data.
• Develop internal standards, patterns, and best practices for building AI/LLM solutions safely within government constraints.

DevOps / MLOps & Knowledge Transfer

• Own the CI/CD and MLOps pipelines for deploying and updating models, APIs, and demo apps (Docker, Kubernetes, GitHub/GitLab, etc.).
• Monitor performance, reliability, and usage of the LLM platform; set up alerting and dashboards.
• Document architectures, runbooks, and how-to guides; train internal developers, architects, and SMEs on how to build on the platform.

Required Qualifications

• Bachelors degree in Computer Science, Engineering, Mathematics, or related field (or equivalent practical experience).
• 5+ yearsof experience in software engineering and/or machine learning engineering.
• 2+ yearsof hands-on experience building or deploying applications using LLMs or other generative AI models.
• Strong proficiency inPythonand one or more modern frameworks for LLM apps (e.g., LangChain, LlamaIndex, OpenAI/Azure OpenAI SDKs, Hugging Face Transformers).
• Experience designing and deploying microservices or APIs (FastAPI, Flask, Node.js, etc.) in cloud environments (Azure, AWS, or GCP).
• Solid understanding of modern data storage and retrieval for AI workloads (SQL/NoSQL databases, vector databases, object storage).
• Familiarity with containerization and orchestration (Docker, Kubernetes) and CI/CD pipelines.
• Demonstrated ability to translate vague requirements into working prototypes and iterate quickly with stakeholders.

Preferred Qualifications

• Experience working withFedRAMP-authorizedenvironments (Azure Government, AWS GovCloud, GCC High) or regulated domains (DoD, DHA, VA, CMS, etc.).
• Understanding of federal security and compliance standards (NIST 800-171, CUI handling, HIPAA, PHI/PII protections).
• Experience implementingRAGarchitectures, multi-step agent workflows, and tool-using LLM agents.
• Hands-on experience with one or more vector databases (Pinecone, Weaviate, Qdrant, pgvector, Redis, etc.).
• Front-end development experience (React, Next.js, or similar) for building interactive demos and user interfaces.
• Prior experience in asmall business / startupor innovation lab environment where you wore multiple hats.
• Prior experience supporting federal proposals, tech volumes, or customer demos/orals is a plus.

Soft Skills & Mindset

• Comfortable working in abuilderrole: you enjoy greenfield problems, ambiguity, and creating structure from scratch.
• Strong communication skills; able to explain AI/LLM concepts and trade-offs to non-technical stakeholders.
• Customer- and mission-focused mindsetexcited about improving government missions with practical, secure AI.
• Collaborative and low-ego; willing to pair with solution architects, SMEs, and BD to co-create demos and offerings.

What Success Looks Like (First 2-3 Months)

• A secure internal LLM environment is stood up and stable (in a FedRAMP or equivalent secure cloud).
• PM Solutions hasreusable demo templates(e.g., knowledge assistant, policy Q&A, document summarization, workflow co-pilot) that can be tailored quickly for pursuits.
• Clear documentation, patterns, and guardrails are in place so other engineers and SMEs can build on the platform.
• Our AI demos meaningfully support proposals and customer conversations, differentiating PM Solutions in the federal AI space.