Employer description & source details
This is a hands-on, forward deployed role for a senior engineer whose first instinct is AI — someone who can shape ambiguous customer problems into precise specifications, author the instructions and prompts that govern agent behavior, build the application and integrations around them, and prove through automated evaluation that the result behaves correctly. This is not a machine learning or AI research role; you will not train foundation models. You own what ships, whether a human or an agent produced it. Build AI-native applications. Design and ship applications built around agentic behavior across Field Service Management — scheduling, dispatch, and workflows that reason over context, automate routine work, and escalate exceptions. Deliver forward deployed. Work embedded in customer environments, tuning instructions and evaluation sets on the ground, using AI to compress turnaround without cutting quality. Author and maintain agentic instructions. Write, structure, and version the system instructions, role definitions, tool descriptions, guardrails, and escalation rules that govern agent behavior — treating them as reviewed, regression-covered engineering artifacts. Build automated evaluation and test non-deterministic behavior. Design and operate golden datasets, multi-turn conversation suites, model-as-judge scoring, CI gates, drift detection, adversarial testing, and grounding checks that make change safe. A demonstrated track record of building, shipping, and operating production software, including hands-on delivery of AI-native application features that real users depend on rather than demos. Direct experience authoring agentic instructions and prompts, designing AI-driven workflows, and building evaluation and testing for non-deterministic behavior is require
Extracted skills: javascript, machine learning