Sr. Full Stack Software Engineer – AI Agents

OVERVIEW

We are seeking a senior full stack software engineer to join our team supporting a customer engagement that builds applications around multi-agent systems developed with Google's Agent Development Kit (ADK) on Google Cloud Platform.

This role is ideal for an engineer who can build the whole application: the agents, the backend services and APIs behind them, and the web interface that users work in. The candidate should have built agents with ADK, deployed them to production on GCP, and operated them afterward, including the evaluation and feedback loops that keep agent behavior reliable over time.

The successful candidate will use AI coding tools, with several agents running at once, as a standard part of daily engineering practice. This engineer will work closely with our technical lead, other engineers, and customer stakeholders to ensure project success from initial definition through final delivery.

REQUIRED CREDENTIALS

• Professional full stack software engineering experience, including production delivery on Google Cloud Platform

• Hands-on experience building and deploying AI agents with Google's Agent Development Kit (ADK)

• Bachelor's degree in Computer Science or equivalent practical experience

REQUIRED QUALIFICATIONS

• ADK agent development, including agent definitions, tool design, and session and state management

• Design and implementation of multi-agent systems, including delegation between agents, sequential, parallel, and loop workflows, and clear boundaries for each agent's context and responsibility

• Feedback loops for agent quality: evaluation sets, automated scoring of agent output, tracing, and using the results to revise prompts, tools, and agent structure

• Frontend development with TypeScript and a modern component framework, including integrating a UI with agent and REST APIs, streaming responses, and authenticated sessions

• Backend service and API development, including relational database design (for example PostgreSQL) with versioned schema migrations

• Authentication and authorization for web applications and agent tool calls, using OAuth 2.0 / OIDC and JWT

• Deployment and operation of agents and application services on GCP using Cloud Run and Agent Engine

• Gemini models on Vertex AI, including prompt design, function calling, and structured output

• Tool integration for agents over MCP and REST APIs

• Daily use of Claude and other AI coding assistants for design, implementation, testing, and review, including running multiple agents concurrently, with the judgment to verify and correct generated output

• Automated testing and CI/CD across the stack, including frontend and end-to-end tests and regression evaluation of agent behavior

USEFUL QUALIFICATIONS

• Agent-to-Agent (A2A) protocol

• Retrieval and grounding with Vertex AI Search or vector search

• Observability for agents and services with Cloud Trace, Cloud Logging, and OpenTelemetry

• Pub/Sub, Memorystore, and BigQuery

• Integration of agents with Gemini Enterprise

• Google Cloud Professional certifications

SCOPE OF WORK

• Design the multi-agent architecture for the solution, including agent roles, delegation, tool boundaries, and state handling

• Implement agents and tools with ADK, including integrations with customer and third-party systems over REST and MCP

• Build the web frontend and the backend services that expose the agents to users, including the user workflows and the APIs they run against

• Deploy agents and application services to GCP on Cloud Run and Agent Engine, with CI/CD and managed environments and releases

• Build and maintain evaluation suites and feedback loops that measure agent output, surface regressions, and drive changes to prompts, tools, and agent structure

• Instrument agents and services with tracing and logging so that behavior in production can be inspected and debugged

• Use Claude and multi-agent workflows to accelerate design, implementation, test generation, and code review across the full stack, with the engineer remaining accountable for the correctness of all delivered code

• Deliver unit, integration, end-to-end, and agent evaluation test coverage, participate in code reviews, and document the architecture and operating procedures for our team and the customer's