Senior Platform Engineer (AI Agentic Engineer)
Posted 22 days ago
The role in short
Embed with a government team to design, integrate, and operationalize AI agents and agentic workflows in secure, classified environments.
Role Description
Defense Unicorns is seeking a senior AI Agentic Engineer to embed with a government team and its technology partners building an emerging AI-powered engineering ecosystem.
This role will focus on designing, integrating, and operationalizing AI agents and agentic workflows within a secure enterprise environment. The engineer will work across large language models, agent frameworks, enterprise data, APIs, developer tooling, and platform services to turn AI capabilities into reliable, mission-relevant workflows.
The ideal candidate is a hands-on engineer who understands both the capabilities and limitations of modern AI systems and can rapidly translate emerging technologies into working solutions. They should be comfortable experimenting with new models and agent frameworks, integrating agents with enterprise systems and tools, evaluating performance and reliability, and working alongside platform engineers and commercial technology partners.
This is not a research-only role. The focus is on moving from AI concepts and prototypes to secure, useful, repeatable agentic capabilities that can operate in production environments.
Responsibilities
- Design, build, and integrate AI agents and multi-agent workflows using modern LLM and agentic technologies.
- Develop agent capabilities that interact with enterprise applications, APIs, data sources, developer tools, and mission systems.
- Integrate models, agents, tools, and enterprise data into secure workflows that can be deployed in controlled and classified environments.
- Work with LLMs, RAG, embeddings, vector databases, tool/function calling, structured outputs, MCP, and related technologies.
- Develop mechanisms for agents to securely discover, access, reason over, and act upon approved enterprise data and services.
- Prototype and rapidly evaluate emerging models, agent frameworks, and AI capabilities to determine where they can create measurable mission or engineering value.
- Develop evaluation and testing frameworks for agent accuracy, reliability, safety, latency, cost, and task completion.
- Implement guardrails and policy controls governing agent behavior, tool access, data access, and human approval points.
- Integrate agentic capabilities with CI/CD and DevSecOps workflows to support repeatable development, testing, deployment, and lifecycle management.
- Collaborate with platform engineers to operationalize AI workloads within Kubernetes/OpenShift environments.
- Work with commercial AI and technology partners to integrate their capabilities into the broader engineering ecosystem.
- Troubleshoot issues across models, agent frameworks, APIs, data sources, networking, identity, and infrastructure.
- Develop reusable patterns for agent deployment, configuration, observability, evaluation, and lifecycle management.
- Document architectures, integration patterns, agent specifications, evaluation results, and operational procedures.
- Identify recurring AI engineering challenges that can be standardized, automated, or productized into reusable capabilities.
- Stay current with rapidly evolving agentic AI technologies and assess their applicability to secure government environments.
Minimum Experience and Qualifications
- Active Top Secret clearance.
- Hands-on experience building applications or systems using large language models and generative AI.
- Experience designing and implementing AI agents or agentic workflows.
- Strong programming experience in Python and/or another modern programming language.
- Experience integrating AI systems with APIs, enterprise applications, databases, and external tools.
- Experience with one or more agent frameworks or orchestration approaches.
- Working knowledge of RAG, embeddings, vector databases, tool/function calling, structured outputs, and prompt engineering.
- Experience developing software in a Git-based, automated CI/CD environment.
- Ability to rapidly prototype, test, troubleshoot, and iterate in ambiguous environments.
- Strong understanding of software engineering fundamentals, including testing, version control, APIs, debugging, and system design.
- Ability to work effectively alongside platform engineers, government personnel, and multiple technology vendors.
Preferred Experience and Qualifications
- Experience with MCP (Model Context Protocol) and tool-based agent architectures.
- Experience with multi-agent systems and agent orchestration.
- Experience with OpenAI, Anthropic, Google, NVIDIA, or comparable foundation model ecosystems.
- Experience with NVIDIA NIM, NeMo, or similar AI inference/model-serving technologies.
- Experience with Red Hat OpenShift AI or Kubernetes-based AI platforms.
- Experience with agent evaluation and observability frameworks.
- Experience implementing AI guardrails, policy enforcement, human-in-the-loop workflows, or AI safety controls.
- Experience with LangGraph, LangChain, LlamaIndex, Semantic Kernel, or comparable frameworks.
- Experience with RAG pipelines, vector databases, knowledge graphs, and enterprise search.
- Experience deploying AI systems in Secret or TS/SCI environments.
- Experience with disconnected or air-gapped AI environments.
- Experience integrating AI into developer/DevSecOps workflows.
- Familiarity with GitLab, Jira, Confluence, Xacta, or similar enterprise engineering and compliance tools.
- Experience working with GPUs, model serving, inference optimization, or AI infrastructure.
- Familiarity with UDS, Zarf, Pepr, Iron Bank, or similar secure software delivery technologies.
- Experience taking AI prototypes into repeatable, production-ready capabilities.
- Ability to evaluate new AI technologies quickly and make pragmatic build/buy/integrate decisions.
What Success Looks Like
- Agentic capabilities move from prototype to mission use quickly, with working agents integrated into real government workflows rather than remaining isolated demos.
- Agents reliably connect to approved enterprise data, tools, and services, with appropriate identity, permissions, guardrails, and human-in-the-loop controls.
- AI capabilities are measurable and trustworthy, with repeatable evaluation methods for accuracy, reliability, task completion, security, and operational performance.
- Reusable agent patterns emerge, reducing the engineering required to build and deploy the next agent or AI-enabled workflow.
- AI integrates cleanly with the broader platform architecture, working effectively across the data, DevSecOps, Kubernetes/OpenShift, model, and application layers rather than creating another technology silo.
- The engineer becomes a trusted AI technical advisor to the Mission Hero, proactively identifying where agentic technology can create mission value—and equally important, where AI is not the right solution.
Prep for this interview
5 questions likely to come up at Defense Unicorns, each matched to a line in this posting.
- 1LLMs, RAG and agents
They wrote: “Design, build, and integrate AI agents and multi-agent workflows using modern LLM and agentic technologies.”
When would you use an agent that takes actions, and how do you keep it safe?
How to answer
Why they ask: Agent work is growing fast in FDE postings. Senior candidates are expected to know where agents fail.
- Use agents for multi-step tasks with clear success checks; avoid them where one wrong action is costly
- Give the agent narrow tools and permissions, not broad access
- Require human approval for risky actions
- Log every step so the customer can audit it
Tip: Start from the cost of a wrong action. Limit what the agent can do, and log everything.
- 2Integrations and APIs
They wrote: “Develop agent capabilities that interact with enterprise applications, APIs, data sources, developer tools, and mission systems.”
Walk me through connecting our product to a customer's internal system that has no documentation.
How to answer
Why they ask: This happens weekly in the job. They want a calm, step-by-step method.
- Find who owns the system and get read access to a test environment
- Watch real traffic or logs to learn the data shapes
- Build a thin adapter with logging around every call
- Agree on how failures get reported, and who fixes what
Tip: Talk about finding the owner of the system first. The code is the easy part.
- 3Cloud, deployment and security
They wrote: “Collaborate with platform engineers to operationalize AI workloads within Kubernetes/OpenShift environments.”
How would you debug a deployment that works in staging but fails at the customer's site?
How to answer
Why they ask: Customer environments differ in ways you can't see. This tests method under pressure.
- Compare versions, config and environment variables
- Check network rules, proxies and certificates (common in enterprise)
- Look at resource limits and permissions
- Add the missing check to your install steps afterwards
Tip: List the usual suspects and how you'd check each without full access.
- 4Coding
They wrote: “Strong programming experience in Python and/or another modern programming language.”
Here's a CSV export from a customer. Write a script that cleans it and loads it into our API.
How to answer
Why they ask: Very close to real FDE work. They watch how you handle bad rows and failures.
- Read with a proper CSV parser; check the header first
- Validate each row and collect errors, not crash on the first one
- Send in batches with retries and a clear report at the end
- Write one or two quick tests for the parsing
Tip: Talk through edge cases out loud: encodings, missing fields, duplicates, rate limits.
- 5Ownership and working with customers
They wrote: “Ability to rapidly prototype, test, troubleshoot, and iterate in ambiguous environments.”
Tell me about a project where the requirements kept changing. What did you do?
How to answer
Why they ask: Ambiguity is the norm. They want proof you can keep moving without a clear spec.
- The project and why requirements kept moving
- What you did to pin them down (short demos, written decisions)
- The trade-offs you made to keep shipping
- The result, with a number if you have one
Tip: Show how you reduced the change, beyond how you coped with it.
Not the one? Get the next one by email
New roles land here every day. Pick what you want to hear about.
Want the job, not just the posting? Learn how FDEs work.
FDE Info is our free weekly newsletter for forward deployed engineers: tools, field notes from real deployments, and how FDE teams hire and pay.
Free. One email a week. Unsubscribe in one click.
