Senior QA Engineer – AI-Native Platform
Posted 31ds ago
Employment Information
Report this job
Job expired or something wrong with this job?
Job Description
Senior QA Engineer building test infrastructure for AI-Native platform at a global OSINT company. Focused on testing non-deterministic AI agent outputs and data isolation.
Responsibilities:
- Build and maintain AI evaluation framework
- Test RAG pipeline quality and retrieval relevance
- Design testing approaches for non-deterministic LLM outputs
- Build multi-tenant isolation test suite, including AI layer isolation
- Write cross-service integration and contract tests
- Test Temporal workflows, event-driven flows, and YAML DSL validation
- Build and maintain automated regression suites
Requirements:
- 3+ years in QA automation / SDET
- Hands-on experience testing AI/ML systems
- Strong Python (pytest, async testing)
- Experience building test infrastructure from scratch
- Experience testing distributed or event-driven systems
- Experience testing multi-tenant applications
- Autonomous work style
- English – B1 level or higher
- RAG / vector search quality evaluation (Nice To Haves)
- LangGraph, LangChain, or similar agent frameworks (Nice To Haves)
- Go test automation (Nice To Haves)
- BentoML / MLflow or equivalent MLOps tooling (Nice To Haves)
- Security testing: prompt injection, context leakage (Nice To Haves)
- Performance / load testing for ML inference endpoints (Nice To Haves)
- Data pipeline testing (Delta Lake, Great Expectations, Dagster) (Nice To Haves)
- Russian – advanced level or higher (Nice To Haves)
Benefits:
- Remote-first setup: work from anywhere in the world (except Russia and Belarus)
- Work on a greenfield, technically challenging product (distributed system at scale)
- Direct collaboration with senior technical stakeholders (engineering leadership, product)
- Ownership and impact: influence QA strategy, tooling, and quality gates from the early stage of the product
- Fast iteration environment: work closely with engineering to improve reliability and release confidence


















