Agentic AI Engineer

Tactical Edge; Washington (District of Columbia); On-site; USD 150,000 - 210,000; Full time

Job summary

Tactical Edge is seeking an engineer to build autonomous AI systems that reason, plan, and execute in production environments. You will design multi-agent architectures, orchestration pipelines, and tool-use frameworks powering enterprise AI solutions.

This is a production-focused role with agents running 24/7 in enterprise settings, requiring strong Python skills, async programming, and a deep understanding of LLM internals.

Qualifications

• Strong Python skills with deep understanding of async patterns and systems design.
• Understands LLM internals (tokenization, attention, context management) beyond API wrappers.
• Production-minded: reliability, observability, and failure modes from day one.
• Comfortable with ambiguity and evolving requirements.
• 3+ years building with LLMs in production environments.
• Experience with agentic frameworks (LangGraph, CrewAI, AutoGen, or custom).
• Strong systems programming skills and familiarity with AWS AI services (Bedrock, SageMaker, ECS).
• Understanding of prompt engineering, RAG architectures, and retrieval systems.

Responsibilities

• Design and build multi-agent systems that operate in production environments.
• Implement tool-use, function calling, and agentic workflows with robust error handling and observability.
• Optimize prompts, manage context windows, and deploy efficient AI pipelines (RAG, retrieval).
• Work with models from Anthropic, OpenAI, and open-source ecosystems.
• Deploy and monitor agents on AWS services with reliability and governance controls.

Skills

Python

Async patterns

Systems design

Agentic frameworks

LLMs in production

Tools

LangGraph

CrewAI

AutoGen

AWS Bedrock

SageMaker

ECS

Job description

We're looking for an engineer who builds autonomous AI systems that reason, plan, and execute in production environments. You'll design multi-agent architectures, orchestration pipelines, and tool-use frameworks that power our enterprise AI solutions.

This is not a research role — this is a production engineering role where your agents run 24/7 in enterprise environments.

Key Responsibilities

• Design and build multi-agent systems using frameworks like LangGraph, CrewAI, or custom orchestration.
• Implement tool-use, function calling, and agentic workflows that operate reliably in production.

LLM Integration & Optimization

• Work with models from Anthropic, OpenAI, and open-source (Llama, Qwen, Mistral).
• Optimize prompts, manage context windows, implement RAG pipelines, and fine-tune models for domain-specific tasks.

Production Systems

• Build agents that run 24/7 in enterprise environments with proper error handling, fallback strategies, human-in-the-loop controls, and observability.
• Deploy on AWS (Bedrock, SageMaker, ECS).

Evaluation & Testing

• Design eval frameworks to measure agent quality, accuracy, and safety.
• Build automated testing for agentic workflows including regression testing and red‑teaming.

Security & Governance

• Implement guardrails, output validation, and audit logging for AI agents operating in regulated industries.
• Self-directed builder who ships without hand-holding.
• Strong Python skills with deep understanding of async patterns and systems design.
• Understands LLM internals (not just API wrappers) — tokenization, attention, context management.
• Production-minded: thinks about reliability, observability, and failure modes from day one.
• Comfortable with ambiguity and evolving requirements.

Preferred Qualifications

• 3+ years building with LLMs in production environments.
• Experience with agentic frameworks (LangGraph, CrewAI, AutoGen, or custom).
• Strong Python and systems programming skills.
• Familiarity with AWS AI services (Bedrock, SageMaker, ECS).
• Understanding of prompt engineering, RAG architectures, and retrieval systems.

How We Work

Outcome-driven

Enterprise-first

Agentic by design

Systems that reason and act safely

Small teams, high ownership

Autonomy with accountability

What You'll Get

• Work on real, production AI deployments
• Enterprise-scale challenges and measurable impact
• Cross-functional collaboration and high ownership
• Flexible work setup where applicable

Hiring Process

1. Intro call

   Fit + context

2. Technical deep dive

   Agent architecture, LLM systems, production design

3. System design exercise

   Multi-agent workflow design

4. Final conversation

   Alignment + next steps

We value clarity, ownership, and thoughtful execution over buzzwords.