AI Agentic Systems Engineer — Top Secret, Onsite DC

Soft Tech Consulting

Washington (District of Columbia)

On-site

USD 120,000 - 165,000

Full time

5 days ago

Be an early applicant

Get more replies from employers

Send a job-specific resume in minutes.

Benefits offered by this job

Medical Insurance

Dental Insurance

Vision Insurance

401K

Short Term Disability

Long Term Disability

Life Insurance

Paid Holidays

PTO

Job summary

Soft Tech Consulting, Inc. is seeking an AI Mid Software Engineer with strong technical acumen to design, implement, and optimize AI-powered solutions built on modern agentic frameworks in Washington, DC.

You will develop intelligent workflows using foundation models and integrate with enterprise data systems. You will collaborate with a team of AI specialists, software engineers, data scientists, solution architects, cybersecurity experts, and product owners to deliver production-ready AI

Qualifications

Bachelor's Degree required (or 9 years of relevant experience).
Minimum 5 years of related experience.
Experience designing and running structured tests for AI behavior.
Proficiency in Python; familiarity with JavaScript or other languages beneficial.
Must be able to communicate effectively, both verbal and written.

Responsibilities

Design, build, and optimize agentic AI systems using orchestration frameworks like LangGraph, CrewAI, Google ADK, or AWS Bedrock AgentCore, etc.
Implement multi-step, tool-enabled AI agentic workflows that combine LLM reasoning, RAG, retrieval systems, external APIs, and custom enterprise tools.
Use leading foundation models (Claude Sonnet, Claude Opus, , etc.) on Amazon Bedrock to build capabilities including structured generation, extraction, summarization, planning, and decision support.
Iteratively tune and refine AI workflows through prompt engineering, instruction design, agent configuration, and tool-use optimization.
Conduct structured testing and evaluation of AI behavior, including regression tests, scenario-based evaluations, synthetic test creation, and model-to-model comparisons.
Diagnose issues such as hallucinations, retrieval mismatches, tool-use errors, or unexpected agent behavior, and implement corrective strategies including guardrails and workflow adjustments.
Integrate AI components with vector databases, graph databases, internal applications, and other infrastructure tools—working alongside engineers who manage the underlying systems.
Document model usage patterns, agent behaviors, experimental findings, workflow designs, and best practices for reproducibility and maintainability.
Collaborate with cybersecurity and governance teams to ensure AI implementations adhere to required policies and safe-use guidelines.