Senior Product Engineer

Agentic Defense Solutions, Inc.; Tysons (VA); On-site; USD 140,000 - 190,000; Full time

Job summary

Agentic Defense Solutions, Inc. is seeking a Senior Product Engineer to own the ARIA product family, the operational trust layer for AI in high-consequence environments.

You will lead AI-enabled product engineering from concept through production, shaping architecture, security, and observability. You will design data models, APIs, and services to support long-running workflows, integrate AI models, and ensure the system remains auditable and governable at scale.

Qualifications

• Experience shipping AI-native software systems.
• Ability to lead end-to-end product engineering from concept to production.
• Strong background in architecture, security, and reliability.

Responsibilities

• Design and build core capabilities across the ARIA platform.
• Own end-to-end capabilities from problem definition to deployment.
• Build and integrate AI agents, tools, and data into workflows.
• Review and improve AI-generated implementations and ensure security.

Skills

AI-native development

API design

Distributed systems

Cloud infrastructure

Code review

Mentoring

Git/GitHub

Tools

Git/GitHub

Docker

Kubernetes

CI/CD

Job description

About the role

Agentic Defense Solutions, Inc. is seeking a Senior Product Engineer to help build the ARIA product family - the operational trust layer for AI in high-consequence environments.

This is an AI-native product engineering role. You will own problems from initial concept through production: understanding the user problem, designing the system, directing AI-assisted implementation, reviewing and validating generated code, and shipping capabilities into real customer environments.

We expect our engineers to use modern AI development tools aggressively. We do not measure engineering contribution by lines of code written by hand. We care about the quality, security, maintainability, and velocity of the systems you ship.

You should have the software engineering depth to build these systems yourself when necessary and, more importantly, to recognize when AI-generated code is wrong, fragile, insecure, over-engineered, or solving the wrong problem.

We are building software at the intersection of AI, distributed systems, data, workflow orchestration, and human decision-making. ARIA enables organizations to put AI into real operational workflows while maintaining control over how AI reasons, what information and tools it can access, when humans must intervene, and how decisions and actions are audited.

That creates interesting engineering problems.

How do you orchestrate long-running agentic workflows that combine models, humans, data, tools, policies, and external systems? How do you represent and reason over complex relationships and mission context? How do you make probabilistic AI systems observable and controllable? How do you build the same platform to operate in modern cloud environments and completely disconnected networks?

You will help us answer those questions.

We're looking for an engineer who enjoys owning difficult problems, moving across traditional engineering boundaries, and shipping products rather than simply completing tickets.

What you'll do

• Design and build core capabilities across the ARIA platform and application ecosystem.
• Own capabilities end-to-end: problem definition, product design, architecture, implementation, testing, deployment, observability, and iteration based on real customer use.
• Build systems that orchestrate AI models, agents, tools, data, policies, human approvals, and external services into reliable operational workflows.
• Develop APIs and services that support complex, long-running, stateful processes.
• Build intuitive user experiences for interacting with AI reasoning, workflows, data, and decisions.
• Design data models capable of representing complex entities, relationships, events, evidence, and operational context.
• Build integrations that allow ARIA to securely interact with customer data sources, applications, models, and infrastructure.
• Help evolve ARIA from an early-stage platform into infrastructure capable of supporting large-scale production deployments.
• Use AI development tools as a core part of how you design, build, test, debug, document, and ship software.
• Effectively direct AI coding agents by decomposing complex engineering problems into clear architectures, specifications, constraints, and implementation plans.
• Review, test, debug, and improve AI-generated implementations rather than treating generated code as inherently correct.
• Use your engineering judgment to identify bad abstractions, security vulnerabilities, architectural problems, unnecessary complexity, performance issues, and subtle correctness failures.
• Develop workflows that allow you to move substantially faster without compromising engineering quality.
• Determine when AI-assisted implementation is appropriate and when a problem requires deeper hands-on engineering.
• Help define how ADS practices AI-native software development as the capabilities of models and development tools continue to evolve.
• Share effective patterns, tooling, and workflows with the broader engineering organization.

We don't care whether you personally typed every line of code. We care whether you understand the systems you ship and are willing to take responsibility for their quality.

AI & Agentic Systems

• Build and improve agentic workflows involving reasoning, planning, tool use, retrieval, memory, and human-in-the-loop decision-making.
• Integrate commercial and open-source foundation models while maintaining a model-agnostic architecture.
• Develop mechanisms for evaluating, testing, and observing AI behavior in production.
• Build systems for managing context, tools, policies, state, memory, and execution across complex AI workflows.
• Help make AI behavior understandable and auditable by capturing the context, evidence, actions, tool calls, policies, and approvals involved in consequential decisions.
• Experiment with emerging models, frameworks, and techniques and determine where they meaningfully improve the product rather than adopting technology for its own sake.
• Help solve the fundamental engineering challenge of combining probabilistic AI reasoning with deterministic operational controls.
• Work directly with product leadership, designers, customers, and mission teams to understand problems before deciding what to build.
• Translate ambiguous operational requirements into simple, reusable product capabilities.
• Move fluidly across frontend, backend, data, AI, infrastructure, and UX when solving a problem requires it.
• Prototype new concepts quickly, test them with users, and turn successful ones into durable product architecture.
• Distinguish between a one-off customer requirement and an underlying capability that should become part of the platform.
• Develop a point of view about the product instead of waiting for fully specified requirements.
• Understand the users and operational environments behind the software you're building.
• Challenge requirements when you believe there is a better way to solve the underlying problem.
• Design scalable services, APIs, and application architectures using modern software engineering patterns.
• Build event-driven and distributed systems that remain understandable, testable, and resilient.
• Design for cloud, hybrid, on-premises, disconnected, and air-gapped deployments.
• Build systems that gracefully handle partial failure, retries, long-running execution, concurrency, and changing external dependencies.
• Improve application performance, reliability, observability, and security as the platform scales.
• Make thoughtful decisions about abstractions and architecture without over-engineering early-stage problems.
• Contribute to technical standards and architectural patterns used across the engineering organization.
• Take responsibility for the quality of what you ship, regardless of whether the implementation was written manually, generated by AI, or some combination of the two.
• Participate actively in architecture discussions, design reviews, and code reviews.
• Raise the technical bar while maintaining the speed required of an early-stage company.
• Help establish engineering patterns, development practices, and tooling as the organization grows.
• Mentor other engineers and share technical knowledge across the team.
• Take ownership when things break and help design systems that make the same failure less likely to happen twice.
• Challenge assumptions — including your own — and advocate for better technical and product decisions.
• Continuously rethink how software should be built as AI changes the economics and mechanics of engineering.

Required Qualifications

• 6+ years of professional software engineering or product engineering experience building and operating production software.
• Deep software engineering fundamentals developed through meaningful experience building real production systems.
• Demonstrated experience designing and building complex applications, platforms, or distributed systems.
• Ability to independently take an ambiguous product or technical problem from initial concept through production implementation.
• Strong proficiency in modern software development and the technical depth to implement complex systems yourself when necessary, even if AI-assisted development is now your default way of working.
• Demonstrated fluency with AI-native software development workflows using tools such as Cursor, Claude Code, Codex, or equivalent agentic development environments.
• Strong experience with modern version control and software delivery practices, including Git/GitHub workflows, pull requests and code review, branching and release strategies, and CI/CD pipelines for reliably testing, building, and deploying production software.
• Ability to effectively direct, review, test, debug, and improve AI-generated implementations.
• Experience designing APIs, services, data models, and production application architecture.
• Experience with modern databases, data infrastructure, and distributed application patterns.
• Experience deploying and operating software in cloud or production environments.
• Strong understanding of testing, observability, performance, reliability, and software security.
• Strong product instincts and a desire to understand why something should be built, not simply how to build it.
• Ability and willingness to move across the stack rather than defining yourself exclusively as a frontend, backend, infrastructure, or AI engineer.
• Ability to communicate technical and product decisions clearly and work effectively with engineers, product leaders, designers, customers, and other stakeholders.
• A track record of shipping meaningful software.

Strongly Preferred

• Experience building products involving LLMs, AI agents, machine learning systems, or AI infrastructure.
• Experience building AI-native products where models and agents are fundamental components of the application rather than isolated features.
• Experience with TypeScript / Node.js and React.
• Experience with Python, particularly for AI, data, or backend services.
• Experience with graph technologies, knowledge graphs, entity/relationship modeling, or graph-based applications.
• Experience building workflow engines, orchestration platforms, distributed task systems, or event-driven architectures.
• Experience with search, retrieval, embeddings, vector databases, or large-scale data processing.
• Experience integrating LLMs through APIs or operating open-source models and inference infrastructure.
• Experience with containers, Kubernetes, cloud infrastructure, and modern CI/CD practices.
• Experience building enterprise software where security, authorization, auditability, and data governance are first-class requirements.
• Experience building developer platforms, APIs, SDKs, or extensible software ecosystems.
• Experience working in a startup or similarly high-ownership engineering environment.

Experience in national security, defense, intelligence, or government is valuable, but it is not required. We care more about your ability to build exceptional products, exercise strong engineering judgment, solve difficult technical problems, and learn a complex mission domain quickly.

The Engineer We're Looking For

You are a builder, but your definition of building has evolved.

You understand that the value of an exceptional engineer is increasingly less about how quickly they can personally type an implementation and more about their ability to understand problems, design systems, make good decisions, direct increasingly capable AI tools, and take responsibility for what ultimately ships.

You have enough traditional engineering experience that you know what good software looks like. You've built systems without an AI agent doing the implementation for you. That experience gives you the judgment to recognize when generated software is elegant, when it's merely functional, and when it's dangerously wrong.

You like understanding the entire system, even when you own only part of it. You can go deep on a difficult backend or distributed-systems problem without losing sight of the user experience that depends on it.

You're comfortable when the requirements aren't completely defined.

Your instinct isn't to ask for a more detailed ticket. It's to understand the problem.

You care about architecture, but you don't confuse complexity with sophistication. You know when an abstraction will make a system dramatically better and when the right answer is simply building the straightforward thing.

You use AI aggressively, but you don't outsource your judgment to it.

You're interested in AI but aren't impressed by demos that only work 80% of the time. You want to understand how AI systems behave when they're placed inside real workflows with real data, real users, real permissions, and real consequences.

You want meaningful ownership.

And you want to ship.

About Agentic Defense Solutions Inc

Agentic Defense Solutions exists to solve the real problem in mission-critical AI: not better models, but the operational trust layer that makes AI governable, auditable, and safe to deploy in high-consequence environments.

Our Mission

Most people think the challenge in mission-critical AI is better models. It isn't. The real problem is there is no control plane for how AI reasons, makes decisions, and executes operations in high-consequence environments.

Agentic Defense Solutions builds the ARIA product family: the operational trust layer for AI where decisions have real-world impact. From the ARIA Platform foundation to the ARIA Hermes application, we provide the infrastructure that makes AI governable, auditable, and safe to deploy so that warfighters, analysts, operators, and decision-makers can leverage AI with confidence.

We believe that AI accountability shouldn't be an afterthought. It should be the foundation. ARIA enforces policy, keeps an immutable audit trail, supports human-in-the-loop, and is built for air-gapped and classified deployment from day one.