Write clean, testable, and resilient production code in Python, C++, or Java, implement automated integration, stress, and regression testing for Large Language Model (LLM) pipelines, proactively identify and pay down technical debt. Deliver complete, robust software components and services that productionalize Data Science prototypes, LLM pipelines, and AI workflows with minimal supervision. Implement orchestration logic, context window management, structured output parsing, and API integrations connecting foundational LLMs to backend systems and frontend interfaces.
Build and maintain automated evaluation pipelines (LLM-as-a-judge, benchmark suites, deterministic unit tests) to monitor hallucination rates, output quality, and inference latency. Partner effectively with Data Scientists, Web Solutions Engineers, and Technical Program Managers to clarify requirements, manage component dependencies, and unblock execution.
Job description provided by Google.
View the original posting
