Skip to main content
ECS logo

Agent Developer

ECS
15 hours ago
Full-time
On-site
Arlington, VA

ECS Federal LLC, the federal arm of Everforth, is seeking an Agent Developer to build production AI agents and workflow-automation capabilities for enterprise and federal customers. This role combines solid software engineering fundamentals with hands-on LLM application development to deliver grounded, testable, secure, and maintainable agent-based solutions.

The developer will work across prompts, tools, APIs, structured outputs, validation, retrieval, and workflow integration. The right candidate is a builder who can move from concept to production while balancing speed, quality, cost, safety, and operational supportability.

The position is onsite in Arlington, VA. ECS Federal LLC is part of Everforth, a $4B global organization with over 10,000 employees. Its federal segment delivers mission-critical technology solutions in data and AI, cybersecurity, and enterprise transformation to defense, intelligence, and federal civilian agencies, supported by nearly 3,500 professionals.

From concept to production: agent development in the field

You design, build, test, deploy, and support production AI agents and AI-assisted workflows. You’ll craft agent workflows for knowledge search, data exploration, SQL generation, pipeline scaffolding, metric design, insight discovery, document processing, and executive narrative generation. You’ll integrate agents with enterprise APIs, data platforms, semantic services, identity systems, and engineering workflows. You’ll implement structured tool calling, schema-constrained outputs, validation logic, retries, exception handling, grounding, and human-in-the-loop controls.

You’ll build retrieval-augmented generation solutions using enterprise search, vector databases, embeddings, chunking, reranking, and citation patterns. You’ll create evaluation frameworks for accuracy, groundedness, safety, latency, cost, regression testing, and failure analysis. You’ll implement observability for prompts, tool calls, model responses, token usage, errors, and workflow performance. You’ll apply secure-development practices, least privilege, secrets management, content filtering, and data-protection controls. You’ll collaborate with architects, data engineers, security teams, domain experts, and delivery leads to move use cases into production. You’ll contribute reusable components, reference patterns, documentation, and AI engineering standards.

What you’ll bring to the project

  • Hands-on experience building LLM-enabled applications, AI agents, or multi-step agentic workflows
  • Strong Python skills and experience developing APIs, services, integrations, and automated tests
  • Experience with tool calling, structured outputs, prompt engineering, context management, and workflow orchestration
  • Experience with retrieval-augmented generation, enterprise search, embeddings, vector databases, and evaluation of grounded outputs
  • Experience integrating enterprise systems through REST APIs, databases, queues, and identity/access services
  • Understanding of AI security concerns including prompt injection, data leakage, unsafe tool execution, model access controls, and auditability
  • Experience with Git, CI/CD, containers, cloud services, monitoring, and production troubleshooting
  • Ability to communicate effectively with technical and nontechnical customer stakeholders
  • Bachelor’s degree in Computer Science, Engineering, Data Science, or a related field, or equivalent professional experience
  • U.S. Citizenship and ability to obtain and maintain the clearance required by the customer; TS/SCI with Poly may be required for Intelligence Community assignments

Nice-to-haves to stand out

  • 5+ years of software engineering experience, including production application development and support
  • Experience with agent frameworks such as LangChain, LangGraph, LlamaIndex, Semantic Kernel, AutoGen, CrewAI, or comparable tools
  • Knowledge of Model Context Protocol, agent registries, tool governance, and secure agent-to-tool interaction
  • Experience with Azure OpenAI, Azure AI Foundry, Amazon Bedrock, Google Vertex AI, OCI Generative AI, OpenAI APIs, or Anthropic APIs
  • Experience with PostgreSQL/pgvector, Elasticsearch, OpenSearch, Pinecone, Weaviate, Milvus, or comparable vector-search technology
  • Familiarity with semantic layers, knowledge graphs, hybrid search, reranking, and enterprise data platforms
  • Experience with Java, Go, JavaScript/TypeScript, or .NET in addition to Python
  • Relevant certifications such as Microsoft Certified: Azure AI Engineer Associate, AWS Certified Machine Learning Engineer – Associate, Google Professional Machine Learning Engineer, Databricks Certified Generative AI Engineer Associate, or Oracle Cloud Infrastructure Generative AI Professional

Where you’ll work and what you’ll need to start

The role is onsite in Arlington, VA and is full-time. You must be a U.S. citizen and able to obtain and maintain the clearance required by the customer; TS/SCI with Poly may be needed for Intelligence Community assignments. ECS Federal LLC is an equal opportunity employer and does not discriminate on any protected characteristic; all qualified applicants will receive consideration for employment without regard to disability, protected veteran status, or any other status protected by applicable law. Everforth ECS is the federal segment of Everforth, a global organization with a strong focus on data and AI, cybersecurity, and enterprise transformation for defense, intelligence, and federal civilian agencies.

Tips for applicants

Lead with a couple of your strongest, relevant projects. Put hands-on LLM-enabled applications and multi-step agent workflows at the top of your resume, so the hiring team sees you can take an idea to production fast.

Show concrete evidence of the must-have skills. Describe your Python-based API work, your experiences with tool calling and structured outputs, and any RAG pipelines or vector-store implementations you’ve built, with measurable outcomes where possible.

Prepare to discuss architecture decisions you’ve made around grounding, validation, retries, and human-in-the-loop controls. Be ready to explain how you approach observability for prompts and tool calls and how you balance latency, cost, and safety in production.

Ask a targeted question about the team’s current data platforms and how they evaluate grounded outputs in production. A good probe is to ask how they measure groundedness and latency in real workflows, and what success looks like in the first 90 days. You might also ask how they balance speed with safety and auditability in evolving agent capabilities.