Job Requirements
Chantilly, VA
Top Secret/SCI Full Scope Polygraph
Mid Level Career (5+ yrs experience)
$200,000 - $265,000
Job Description
AI/ML Engineer (LLM and Agent Applications), TS/SCI Full Scope Poly
Waveguide | Onsite, Washington DC metro | Full time | 2 openings | Pay: DOE
Waveguide is hiring two AI/ML engineers to build mission AI applications onsite at a government sponsor's headquarters in the Washington DC metro. You'll build production LLM applications: agent workflows, retrieval-augmented generation (RAG), and the Python services and data underneath them. This is hands-on engineering for analysts who will rely on what you ship.
What you'll do
- Build and maintain AI/ML applications centered on agent workflows, LLM integration and RAG.
- Design agent architectures: orchestration logic, tool use, memory and context handling, and connections to enterprise tools and data sources.
- Build RAG pipelines end to end, from document processing and embeddings to retrieval tuning, prompt assembly and response quality.
- Write and maintain Python REST APIs, backend services, user-facing features and data workflows.
- Model structured and unstructured data, refine Postgres schemas, benchmark models in MLflow or similar tools, and tune for latency, reliability and cost.
What you need
- An active TS/SCI with a full scope polygraph. This is required on day one.
- Hands-on experience building LLM-enabled applications with agents, RAG or similar retrieval.
- Strong Python, including LLM application development with orchestration, tool use, context management and outside integrations.
- A solid grasp of the fundamentals: transformers, how LLMs behave, data processing and integration patterns.
- REST API work, ideally Python services, plus PostgreSQL or another relational database, including schema design, indexing and query tuning.
- Production habits: testing, reliability, deployment and performance.
- Experience at one of three levels:
- Full Performance: a bachelor's and 3 years, a master's and 1 year, or a PhD.
- Senior: a bachelor's and 8 years, a master's and 7, or a PhD and 6.
- Expert: a bachelor's and 12 years, a master's and 10, or a PhD and 8.
Nice to have
- MCP (the protocol that connects AI agents to tools and data) or something like it.
- Full-stack AI apps, from frontend to backend to orchestration.
- FastAPI or Flask, Docker, and AWS or another cloud.
- LLM memory management, evaluation, observability, context window management, and cost and performance tuning.
- MLflow, vector databases, embedding models and semantic search.
- Geospatial AI/ML work with PyTorch, TensorFlow, scikit-learn or similar, including building training and test datasets and reporting model performance.
Schedule: Monday through Friday, 40 hours, onsite. You'll need to be in during the sponsor's core hours of 10 AM to 2 PM, with the occasional weekend in a crunch. There's no travel.
Waveguide | Onsite, Washington DC metro | Full time | 2 openings | Pay: DOE
Waveguide is hiring two AI/ML engineers to build mission AI applications onsite at a government sponsor's headquarters in the Washington DC metro. You'll build production LLM applications: agent workflows, retrieval-augmented generation (RAG), and the Python services and data underneath them. This is hands-on engineering for analysts who will rely on what you ship.
What you'll do
- Build and maintain AI/ML applications centered on agent workflows, LLM integration and RAG.
- Design agent architectures: orchestration logic, tool use, memory and context handling, and connections to enterprise tools and data sources.
- Build RAG pipelines end to end, from document processing and embeddings to retrieval tuning, prompt assembly and response quality.
- Write and maintain Python REST APIs, backend services, user-facing features and data workflows.
- Model structured and unstructured data, refine Postgres schemas, benchmark models in MLflow or similar tools, and tune for latency, reliability and cost.
What you need
- An active TS/SCI with a full scope polygraph. This is required on day one.
- Hands-on experience building LLM-enabled applications with agents, RAG or similar retrieval.
- Strong Python, including LLM application development with orchestration, tool use, context management and outside integrations.
- A solid grasp of the fundamentals: transformers, how LLMs behave, data processing and integration patterns.
- REST API work, ideally Python services, plus PostgreSQL or another relational database, including schema design, indexing and query tuning.
- Production habits: testing, reliability, deployment and performance.
- Experience at one of three levels:
- Full Performance: a bachelor's and 3 years, a master's and 1 year, or a PhD.
- Senior: a bachelor's and 8 years, a master's and 7, or a PhD and 6.
- Expert: a bachelor's and 12 years, a master's and 10, or a PhD and 8.
Nice to have
- MCP (the protocol that connects AI agents to tools and data) or something like it.
- Full-stack AI apps, from frontend to backend to orchestration.
- FastAPI or Flask, Docker, and AWS or another cloud.
- LLM memory management, evaluation, observability, context window management, and cost and performance tuning.
- MLflow, vector databases, embedding models and semantic search.
- Geospatial AI/ML work with PyTorch, TensorFlow, scikit-learn or similar, including building training and test datasets and reporting model performance.
Schedule: Monday through Friday, 40 hours, onsite. You'll need to be in during the sponsor's core hours of 10 AM to 2 PM, with the occasional weekend in a crunch. There's no travel.
group id: 91173751