Job Details

Primary Skills

Agentic Engineer GCP

    •     Build ADK agents, tools and sub-agents to the pod lead's design. This includes state, sessions and Memory Bank where the workflow needs memory across sessions.

    •     Integrate APIs and BigQuery views, reaching data through curated datasets and never raw PII tables.

    •     Write unit tests, tool contract tests and eval cases. Run user simulation before each release.

    •     Add guardrails: Model Armor policies, confirmation gates on any action that publishes, spends or changes customer-facing content.

    •     Instrument tracing and logging so every tool call can be traced to an agent identity.

Must have

    •     4+ years in Python or TypeScript backend engineering.

    •     1+ year building LLM applications with tool calling. ADK experience preferred; Gen AI SDK or LangGraph acceptable with an ADK ramp.

    •     GCP fundamentals: Cloud Run, BigQuery, IAM, CI/CD.

    •     Has written eval or test harnesses for non-deterministic output.

Nice to have

    •     Google Cloud Professional ML Engineer or Cloud Developer certification.

    •     Experience with marketing APIs: Google Ads, DV360, CM360, GA4 or Merchant Center.

    •     Prompt and instruction versioning practice.

QA-focused seats additionally need: experience building golden datasets, LLM-as-judge setups with human calibration, and regression gates.