AI Platform Engineer

Job summary

The AI Platform Engineer builds and operates production LLM agent systems and anomaly detection tools, owning delivery from architecture and implementation through deployment and monitoring. The role develops MCP servers and REST API integrations, deploys containerized services through CI/CD to Kubernetes or cloud infrastructure, and improves time-series forecasting and alert thresholds using production metrics. It also applies enterprise AI safeguards, including access controls, PII and secrets handling, audit logging, cost limits, and model evaluation.

Qualifications

• At least 4 years of experience shipping LLM agent systems to production.
• End-to-end production agent experience spanning architecture, implementation, CI/CD deployment to Kubernetes or cloud, and monitoring.
• Experience building MCP server tools and integrating REST APIs.
• Understanding of production agent issues, including timeouts, oversized responses, retries, and token cost management.
• Hands-on experience with time-series forecasting or anomaly detection on production metrics.
• Experience tuning alert thresholds against real incidents and balancing false positives with missed alerts.
• Regular use of AI coding tools and disciplined review and validation of generated code.
• Practical experience applying enterprise AI safeguards, including code-enforced access controls, PII and secrets handling, audit logging, cost limits, and model evaluation.
• Comfort working through security and compliance reviews.
• Bachelor’s degree in Computer Science, Engineering, or a related field, or equivalent experience.

Responsibilities

• Design, build, and operate production LLM agent systems using Google ADK or comparable frameworks.
• Own end-to-end delivery, including architecture, coding, containerization, CI/CD, deployment, and post-launch monitoring.
• Build agent tools as MCP servers and integrate them with REST APIs.
• Manage production concerns such as tool timeouts, oversized responses, retries, and token costs.
• Develop or improve time-series forecasting and anomaly detection using production metrics.
• Tune detection thresholds against real incidents and balance false positives against missed alerts.
• Extend AI coding tools for team use through plugins, skills, slash commands, hooks, or MCP integrations.
• Apply AI safeguards, including access controls, PII and secrets handling, audit logging, cost limits, and model output evaluation.
• Identify operational needs and prototype tools that engineering and operations teams adopt.

Skills

Degrees

Industry

Relocation