Senior Agentic AI Engineer @ Houston, TX :: FTE
Job details
- Pay
- $150,000 – $160,000 a year
- Work mode
- On-site
- Employment
- Full-time
- Level
- Senior
- Experience
- 8+ years
- Posted
- Oct 8, 2026
- Last confirmed open
- Oct 8, 2026
About this role
Role: Senior Agentic AI Engineer
Location: Houston, TX(Onsite)
Exp: 8+ years
$150k to $160k
Full Time
Must-Have
- Software engineering experience, including 2+ hands-on experience building LLM-based applications in production.
Strong Python (TypeScript a plus); deep familiarity with LLM APIs (OpenAI, Anthropic, Gemini, open-weight models) and prompt/context engineering.
Hands-on experience with agent frameworks, tool/function calling, structured outputs, and multi-agent orchestration patterns.
Experience with vector databases (Pinecone, Weaviate, pgvector, etc.), embeddings, and retrieval system design.
Solid backend fundamentals
APIs, async systems, distributed services, Docker/Kubernetes, and cloud platforms.
Practical knowledge of LLM evaluation, safety, and reliability techniques; comfort working with imperfect, probabilistic systems.
Excellent communication skills and a track record of ownership in fast-moving environments.
Good-to-Have
Fine-tuning / RLHF experience, open-source contributions to AI tooling, or published work on agents.
Experience with MLOps/LLMOps platforms (LangSmith, Langfuse, Weights & Biases, MLflow) and enterprise security/compliance requirements.
Architect and build multi-step, tool-using LLM agents (planning, memory, retrieval, orchestration) for production use cases.
Design robust agent workflows using frameworks such as LangGraph, CrewAI, AutoGen, or the OpenAI/Anthropic agent SDKs; integrate tools via MCP and function calling.
Collaborate with solution architects, product teams, and application owners to implement integration designs.
Build RAG pipelines, vector search, and context-management strategies that keep agents accurate and cost-efficient.
Strong Python (TypeScript a plus); deep familiarity with LLM APIs (OpenAI, Anthropic, Gemini, open-weight models) and prompt/context engineering.