| Location | Austin, TX |
I. DESCRIPTION OF SERVICES
Researching, designing, implementing and managing software programs. Testing and evaluating new programs. Working closely with other developers, UX designers, business and systems analysts.
AI Agent Engineer Designs and develops AI-driven agentic solutions, including autonomous workflows and Retrieval-Augmented Generation (RAG) systems, to enhance productivity, automate processes, and support intelligent decision-making with a focus on governance, security, and cost efficiency.
II. CANDIDATE SKILLS AND QUALIFICATIONS
Minimum Requirements:
Candidates that do not meet or exceed the minimum stated requirements (skills/experience) will be displayed to customers but may not be chosen for this opportunity.
Years
Required/Preferred
Experience
4
Required
experience in AI/ML engineering or advanced data science
4
Required
Proven track record of building and deploying production-grade autonomous agents.
4
Required
Strong experience in context engineering
4
Required
Deep experience with LangChain, LangGraph, CrewAI, or AutoGPT.
4
Required
Experience implementing RAG architectures using vector databases
4
Required
Proficiency in Python and AI/ML libraries (OpenAI, Hugging Face, Azure AI)
4
Required
Experience integrating LLMs via APIs Knowledge of AI governance, model lifecycle management, and evaluation
4
Required
Experience implementing and extending the Model Context Protocol (MCP) to provide LLMs with secure, standardized access to local and remote data sources Experience implementing AI guardrails, content filtering, and safety controls
4
Required
Understanding of data privacy and handling of sensitive data (PII/PHI)
2
Preferred
Experience building multi-agent or autonomous agentic workflows
2
Preferred
Experience optimizing LLM cost, token usage, and performance
2
Preferred
Familiarity with enterprise AI deployment patterns and scalability considerations
Solicitation Reference Number: 529601688
Customer/Dept: Texas Health and Human Services Commission
Hybrid - 3 days remote with 2 days (Tues & Wed) required to be onsite