AI Engineer

LTM is hiring an AI Engineer to design and deliver enterprise agentic AI systems with a focus on scalable architecture and secure integration. The role centers on building autonomous multiagent workflows and governing their deployment end to end within a strict CI/CD environment on Google Cloud.

Based in Southlake, TX (hybrid, 3 days), this position is suited for an experienced engineer who can turn complex agent orchestration into reliable production-grade applications, while actively monitoring performance and controlling token and cost spend.

Responsibilities

• Design and build production-grade Python applications using orchestration frameworks such as LangChain, LangGraph, and CrewAI for autonomous multiagent systems, including stateful reasoning and complex RAG (Retrieval-Augmented Generation) workflows.
• Deploy, scale, and govern AI agents using the Google Cloud Gemini Enterprise Agent Platform through the Vertex AI Agent Builder ecosystem, including Agent Development Kit (ADK) code-first deployment, Agent Engine stateful runtimes, and Agent Studio for prototyping.
• Implement secure inter-agent collaboration and data access by using Agent2Agent (A2A) and the Model Context Protocol (MCP), connecting agentic workflows to enterprise databases, remote MCP servers, and third-party workflow APIs.
• Monitor agent behavior and optimize LLM prompt structures, context windows, and caching mechanisms to improve reasoning efficiency while minimizing token usage and operational costs.
• Design and maintain hybrid search using Vertex AI Vector Search and Vertex AI Search to ground decisions in authoritative enterprise data, including local files and external specialized data sources.
• Architect decentralized agent systems under zero-trust principles, enforcing data privacy protections with Vertex AI Model Armor to prevent prompt injections and managing permissions securely through Agent Identity and Google Cloud IAM.
• Build and maintain robust CI/CD pipelines to automate testing, versioning, and deployment of agentic systems, using Vertex AI Agent Engine runtime tracing, logging, and Unified Trace Viewers to debug complex reasoning loops in production.

Requirements

• Strong mastery of Python and standard enterprise software design patterns.
• Deep hands-on experience building complex, production-ready agentic workflows with LangChain, LangGraph, CrewAI, or Google’s native Agent Development Kit (ADK).
• Proven expertise with Gemini Enterprise Agent Platform and Vertex AI Agent Builder, including GKE (Google Kubernetes Engine), Cloud Run, and IAM.
• Familiarity with modern agent communication standards such as MCP and A2A.
• Strong understanding of Vertex AI Vector Search or standalone vector databases, including semantic search, metadata filtering, and embedding lifecycle management.
• Proficiency with modern CI/CD tools such as GitHub Actions, GitLab CI, and containerization using Docker and Kubernetes (Docker/Kubernetes).
• 5 years of software engineering experience, with at least 2/3 years dedicated to building and deploying AI/LLM-powered applications and agentic systems at enterprise scale.
• Bachelor’s or Master’s degree in Computer Science, AI, Data Science, or a related technical field, or equivalent practical experience.

Technologies

• Python
• LangChain, LangGraph, CrewAI
• Gemini Enterprise Agent Platform, Vertex AI, Vertex AI Agent Builder
• Agent Development Kit (ADK), Agent Engine, Agent Studio
• Agent2Agent (A2A) protocol, Model Context Protocol (MCP)
• Vertex AI Vector Search, Vertex AI Search, Vertex AI Model Armor
• Agent Identity, Google Cloud IAM
• Vertex AI Agent Engine Runtime, Unified Trace Viewers
• GitHub Actions, GitLab CI, Docker, Kubernetes (GKE)
• Cloud Run

Benefits

• Comprehensive Medical Plan (Medical, Dental, Vision)
• Short Term and Long-Term Disability Coverage
• 401(k) Plan with Company match
• Life Insurance
• Vacation Time, Sick Leave, Paid Holidays
• Paid Paternity and Maternity Leave

Employment type: Full-Time

Compensation range: USD 107,720 - 112,720 per year

Location: Southlake, TX (hybrid, 3 days)

Mandatory skills: MLOPS, Python