AI Architect (Technical Lead)

Expression•Shirlington, VA

2d•$200,000 - $230,000•Onsite

About The Position

Expression is seeking an experienced AI Architect to serve as Technical Lead designing and technically integrating advanced Artificial Intelligence (AI) and Large Language Model (LLM) capabilities supporting the Department of Homeland Security (DHS). The role supports the development and integration of secure, mission-focused AI agents within the DHS AI environment, adapting multi-agent capabilities such as workflow orchestration, retrieval-augmented generation (RAG), model evaluation, and governed enterprise data integration to enhance investigative analysis, information retrieval, report generation, and human decision support. The successful candidate will be a hands-on AI architect and technical leader who defines system architecture, develops and reviews production-quality software, integrates LLMs with enterprise data systems, and mentors engineers. The position requires expertise in modern AI engineering, distributed application architectures, secure API development, and responsible AI deployment in highly regulated environments. The role works closely with AI engineers, data engineers, program leadership, and Government stakeholders to deliver reliable, explainable, secure, and scalable AI capabilities. A Top Secret/SCI security clearance is required. The position is located in Shirlington, Virginia.

Requirements

Bachelor's degree in Computer Science, Software Engineering, Computer Engineering, Data Science, Artificial Intelligence, or a related technical field, or equivalent relevant professional experience.
12+ years of professional experience in software engineering, AI engineering, data engineering, or enterprise application development, including 3+ years in software architecture or technical leadership roles.
Demonstrated hands-on experience architecting, developing, and deploying enterprise LLM-powered and generative AI applications, including agentic AI workflows, multi-agent orchestration, and RAG.
Experience designing and implementing multi-agent AI systems using LangGraph, LangChain, Semantic Kernel, AutoGen, or equivalent technologies, including Responder–Evaluator–Reporter workflows, model comparison, and AI output verification.
Experience with LLM prompt engineering, prompt version management, structured generation, tool/function calling, agent execution controls, and human-in-the-loop workflows.
Strong Python software engineering skills, including RESTful API development, asynchronous programming, modular application design, and modern software development practices.
Experience designing distributed applications, microservices, or event-driven architectures and integrating AI capabilities with structured and unstructured enterprise data through secure APIs and retrieval architectures.
Experience implementing secure, permission-aware enterprise data integrations, including authentication, authorization, identity propagation, data governance, source traceability, and auditability.
Experience designing and evaluating reliable, explainable AI systems, including factual grounding, source citations, hallucination mitigation, model performance evaluation, and responsible human oversight.
Experience with LLMOps and AI assurance practices, including automated regression testing, prompt-injection defenses, model/prompt version tracking, application observability, and performance monitoring.
Experience with cloud-native architectures and containerized application deployment using Docker, AWS or Azure, Kubernetes, Terraform, Helm, or comparable technologies.
Experience implementing CI/CD pipelines, automated testing, version control, and software configuration management to support secure, reliable application delivery.
Demonstrated ability to translate technical and mission requirements into scalable system architectures, implementation plans, interface specifications, and engineering deliverables.
Experience leading technical design and code reviews, troubleshooting complex software integrations, mentoring engineers, and providing hands-on technical guidance throughout the software development lifecycle.
Strong written and verbal communication skills, including experience producing technical documentation, presenting architectural decisions, and supporting technical demonstrations, Government acceptance testing, and operational pilot deployments.
Top Secret/SCI security clearance required.

Nice To Haves

Advanced degree in Computer Science, Artificial Intelligence, Machine Learning, Software Engineering, or a related technical discipline.
Experience integrating enterprise AI applications with Palantir, Databricks, Elasticsearch, or similar governed data and analytics platforms.
Experience developing Python backend services using FastAPI, Pydantic, asyncio, or comparable frameworks.
Experience supporting secure software development and deployment in Federal Government, Department of Defense, Intelligence Community, or Department of Homeland Security environments.
Familiarity with Government cybersecurity, privacy, accessibility, and software authorization processes.
Experience adapting existing AI research, algorithms, prototypes, or software components into operational enterprise applications.
Experience supporting AI-enabled document processing, intelligent search, case management, investigative analytics, or automated report generation.

Responsibilities

Lead the technical architecture, design, and implementation of enterprise AI and agentic AI solutions within the Government's existing AI environment, emphasizing reuse of approved infrastructure, services, and model endpoints.
Define scalable, secure, and maintainable software architectures integrating LLMs, specialized AI agents, enterprise data services, APIs, and human-in-the-loop workflows.
Design and implement multi-agent orchestration frameworks, including Responder–Evaluator–Reporter workflows, model selection, prompt management, tool invocation, and structured output validation.
Lead integration of Government-approved frontier AI models through authorized endpoints, including model configuration, version management, performance evaluation, and controlled execution.
Architect RAG solutions integrating authorized enterprise data sources while preserving source attribution, access controls, data lineage, and traceability.
Guide integration with Palantir, Databricks, Elasticsearch, and other approved Government data services through secure APIs and governed interfaces.
Design secure AI agent execution frameworks with identity propagation, role- and purpose-based access controls, bounded tool permissions, approval checkpoints, execution limits, and safe failure mechanisms.
Develop and review production-quality backend services, APIs, and orchestration components using Python and modern software engineering practices.
Lead architecture and code reviews, resolve complex integration challenges, and establish engineering standards for maintainability, reliability, security, and performance.
Design evaluation and testing approaches for LLM and multi-agent applications, including factual accuracy, consistency, explainability, latency, scalability, and computational cost.
Establish observability, logging, prompt and model version tracking, and auditability across AI workflows so outputs can be traced to supporting sources and reviewed by authorized personnel.
Collaborate with cybersecurity, DevSecOps, privacy, and Government stakeholders to align system designs with approved deployment pipelines, security controls, authorization requirements, and release processes.
Support adaptation of existing AI algorithms, prompts, and software components into Government-approved investigative workflows, documenting technical lineage, modifications, and integration requirements.
Produce solution architecture documents, system design specifications, interface control documents, data-flow diagrams, technical evaluations, and engineering implementation guidance.
Support technical demonstrations, prototype evaluations, user acceptance activities, and Government design reviews.
Provide technical leadership and mentorship to AI and data engineers while maintaining hands-on involvement in architecture, development, debugging, and integration.
Evaluate proposed third-party AI tools, frameworks, and software components for technical suitability, security implications, integration complexity, and reuse of existing Government capabilities, coordinating recommendations through established approval processes.