V
Agentic Software Developer in Test (SDET)
Job Description
Agentic Software Developer in Test (SDET)
Core Focus Automation Solutions for API Fuzzing & Multi-Agent Validation
API Fuzzing Build and scale automated fuzzing solutions to maximize contract and security coverage.
Spec Automation Automate document and spec generation to align testing frameworks with active schemas.
CI/CD Pipelines Partner with DevOps to embed automated, fail-fast evaluation gates directly into build systems.
Auditing & Compliance Implement comprehensive logging tracking system prompts and outputs for verification.
AI Evaluation & Architecture Promptfoo, Ragas, A2A (Agent-to-Agent) Testing, Cloud Run
Scenario & Assertion Targets
Engineering Action Items
Epic 1 Core Agent Behavior & Guardrails
Validate prompt injection resiliency, hallucination boundaries, and precise JSON schema tool arguments.
Configure localized Promptfoo/Ragas test beds on Cloud Run seeded with 100+ adversarial vectors.
Epic 2 Multi-Agent Mesh & Orchestration
Ensure downstream cascade failure logging, infinite loop detection, and cross-event state consistency.
Construct an autonomous agent mocking framework and loop-checking logic in the configuration registry.
Epic 3 Pipelines & Observability
Build fail-fast pull request blockers based on semantic scores and 60-second canary rollbacks.
Integrate assertion steps in GitHub Actions / Cloud Build; map token and telemetry joins to dashboards.
Core Focus Automation Solutions for API Fuzzing & Multi-Agent Validation
- Role Objective & Key Responsibilities
API Fuzzing Build and scale automated fuzzing solutions to maximize contract and security coverage.
Spec Automation Automate document and spec generation to align testing frameworks with active schemas.
CI/CD Pipelines Partner with DevOps to embed automated, fail-fast evaluation gates directly into build systems.
Auditing & Compliance Implement comprehensive logging tracking system prompts and outputs for verification.
- Required Technical Stack
AI Evaluation & Architecture Promptfoo, Ragas, A2A (Agent-to-Agent) Testing, Cloud Run
- Core Testing Framework Blueprint
Scenario & Assertion Targets
Engineering Action Items
Epic 1 Core Agent Behavior & Guardrails
Validate prompt injection resiliency, hallucination boundaries, and precise JSON schema tool arguments.
Configure localized Promptfoo/Ragas test beds on Cloud Run seeded with 100+ adversarial vectors.
Epic 2 Multi-Agent Mesh & Orchestration
Ensure downstream cascade failure logging, infinite loop detection, and cross-event state consistency.
Construct an autonomous agent mocking framework and loop-checking logic in the configuration registry.
Epic 3 Pipelines & Observability
Build fail-fast pull request blockers based on semantic scores and 60-second canary rollbacks.
Integrate assertion steps in GitHub Actions / Cloud Build; map token and telemetry joins to dashboards.
More Info
Key Skills
Agent-to-Agent Testing
Cloud Run
API Fuzzing
Auditing Compliance
Ragas
Promptfoo
