AI Software Engineer, Applications and Agents

General Dynamics Mission Systems, Inc•Manassas, VA
•$156,965 - $170,000•Hybrid

About The Position

The AI Software Engineer, Applications and Agents designs, codes, tests, evaluates, secures, deploys, and improves full-stack AI products. The engineer works directly with users, domain experts, product owners, and technical teams to take problems from definition through working products and measurable outcomes. This hands-on individual-contributor role combines software engineering, enterprise data, and AI to build secure applications and re-engineer operations using approved cloud services, models, APIs, data, identity, security, and scalable application patterns across Azure and AWS. The role involves building production AI applications, agents, APIs, services, integrations, and enterprise workflows; developing agentic workflows and full-stack products; and owning delivery from architecture and implementation through release, adoption, support, and lifecycle improvement. It also includes establishing quality, evaluation, and safety practices; building automated evaluation and testing pipelines; implementing security controls; and instrumenting products for observability and operational support. The engineer partners with business and mission teams to identify appropriate AI and non-AI solutions, improve operational processes, develop reusable components, and measure adoption and impact. The role provides advanced technical direction through engineering judgment, architecture and code reviews, mentoring, and technical work direction, without requiring people-management authority. The position is hybrid near a major GDMS facility preferred; remote work may be considered for an exceptional candidate able to support periodic on-site collaboration.

Requirements

  • Bachelor's degree in Software Engineering or a related Science, Technology, Engineering, or Mathematics field plus at least 8 years of relevant experience; or a master's degree plus at least 6 years of relevant experience.
  • U.S. citizenship is required due to the nature of work performed within the facilities.
  • Advanced or expert proficiency in coding, testing, integrating, documenting, deploying, and operating production software products.
  • Strong production coding ability in Python and at least one additional language such as TypeScript, JavaScript, Java, Go, C++, C#, or Rust.
  • Experience delivering full-stack or backend cloud applications using APIs, data stores, eventing, authentication, authorization, testing, CI/CD, telemetry, and production support.
  • Hands-on experience building generative AI applications, agents, RAG solutions, tool-using workflows, or comparable intelligent systems in production.
  • Experience with prompt and context engineering, retrieval and grounding, memory, tool calling, orchestration, human-in-the-loop design, and agent evaluation.
  • Experience integrating enterprise data, APIs, business systems, identity, permissions, and secure tool execution into production applications.
  • Experience building automated tests and evaluations for functional behavior, model and agent quality, safety, reliability, latency, and regressions.
  • Production experience integrating enterprise data and cloud-native AI services across Azure and AWS, including model APIs, orchestration, storage, eventing, identity, security, telemetry, and modern containerized or serverless application patterns.
  • Ability to solve ambiguous problems, work directly with technical and nontechnical users, and own delivery through production and measurable outcomes.
  • Written and verbal communication skills, including the ability to explain architecture, model behavior, risk, tradeoffs, and product evidence to technical and executive audiences.

Responsibilities

  • Design, write, test, integrate, review, deploy, and operate production AI applications, agents, APIs, services, and user experiences.
  • Work with business and engineering teams to build production-scale AI applications, integrations, and end-to-end enterprise workflows.
  • Build agentic workflows using planning, reasoning, context management, retrieval, tool use, human approval, and multi-agent collaboration where appropriate.
  • Integrate approved commercial, open-weight, self-hosted, and GDMS-controlled models through secure interfaces and runtime services.
  • Develop full-stack products using modern frontend, backend, API, event-driven, data, and cloud-native engineering patterns.
  • Own delivery from problem definition through architecture, implementation, testing, production release, adoption, support, and lifecycle improvement.
  • Build secure enterprise integrations with business systems, engineering tools, data services, identity, authorization, and workflow platforms.
  • Define product-quality, task-success, groundedness, safety, latency, reliability, and consumption criteria with product owners and AI Reliability Engineering.
  • Build automated evaluation pipelines, representative test suites, golden datasets, regression and integration tests, automated raters, and human-review workflows.
  • Evaluate prompts, instructions, context, retrieval, tools, model configurations, routing, and fallback behavior against approved use cases.
  • Implement controls for hallucination, prompt injection, data leakage, unsafe tool use, excessive autonomy, cross-tenant access, and policy violations.
  • Use red teaming, failure testing, observability, production feedback, and incident learning to improve application and agent behavior.
  • Use approved CI/CD, infrastructure-as-code, release, artifact, security, and evidence workflows to deliver supported production products.
  • Build authentication, authorization, secrets, audit, data-protection, approval, stop, rollback, and recovery behavior into product design.
  • Instrument metrics, logs, traces, quality signals, agent actions, model calls, tool use, latency, consumption, failures, and user outcomes.
  • Troubleshoot complex failures across application, model, retrieval, tool, data, identity, network, and platform layers, and implement permanent corrections.
  • Convert defects, incidents, and user feedback into regression tests, permanent engineering corrections, and reusable platform improvements.
  • Partner with functions, programs, engineers, and domain experts to decompose work, identify suitable AI and non-AI solutions, and define measurable outcomes.
  • Re-engineer operational processes to reduce manual effort, improve flow and controls, and create scalable workflows combining people, automation, and AI.
  • Develop prototypes when needed to resolve uncertainty, then productionize, transfer, or stop the work based on evidence.
  • Translate recurring product and workflow needs into reusable components, reference applications, templates, and integrations.
  • Measure adoption, task success, quality, cycle time, productivity, reliability, and mission or business impact after release.
  • Define scope, architecture, engineering evidence, schedule, risks, and acceptance expectations for personally executed or technically led work.
  • Provide code and architecture reviews, pairing, mentoring, and technical work direction without requiring people-management authority.
  • Use AI coding agents and engineering assistants for implementation, testing, review, diagnosis, documentation, and delivery while maintaining verification and accountability.
  • Keep current with generative AI, agent, software-engineering, evaluation, cloud, security, and product-delivery practices, and introduce validated improvements into GDMS engineering.

Benefits

  • Highly competitive benefits
  • Flexible work environment
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service