AI Reliability Engineer

Palona AI•Seattle, WA

1d•Remote

About The Position

Palona brings the newest advancements in generative, multimodal AI, Real World AI, to the hospitality industry. Our engineers deliver at a fast pace to stay ahead of the AI industry, employing generative AI models to power products that are constantly evolving. In this environment, traditional software excellence principles must adapt to probabilistic outputs -- AI systems don't fail the way conventional software does. This role exists to build the engineering discipline that catches failures before customers do. You will work across evaluation pipelines, observability, cloud infrastructure, and CI/CD for Palona's AI agent platform. The role is equal parts DevOps and AI reliability: managing production infrastructure while building the tooling that keeps AI agents performing at quality.

Requirements

Minimum 2 years of professional software engineering experience
Proficiency in Python
Experience with cloud platforms (AWS, Azure, or GCP)
Experience with monitoring and observability tools (Datadog, CloudWatch, Grafana, or similar)
Familiarity with CI/CD pipelines and infrastructure as code
Experience with APIs and distributed systems
Willingness to learn new AI/LLM concepts, frameworks, and technologies

Nice To Haves

Experience writing test frameworks or automated evaluation systems
Exposure to LLMs, prompt engineering, or AI agent systems
Startup experience, or ability to thrive in fast-paced environments
Background in NLP, computer vision, or AI agent systems

Responsibilities

Instrument services, build dashboards, and maintain alerts that surface quality degradation, latency issues, and system anomalies in production
Write and maintain automated tests for agent output quality, including deterministic checks and LLM-as-judge evaluations
Build and maintain deployment pipelines with quality validation gates, automating release workflows across multiple environments
Manage and improve cloud resources and maintain infrastructure as code across environments, optimizing for cost, reliability, and scalability
Build and extend evaluation pipelines that assess AI agent conversation quality, accuracy, and safety, collaborating with product and engineering to evolve evaluation criteria