SDET - AI ML Validation Engineer

Experience Range: 7-10 years

Job Location: Bangalore with Hybrid Primary Skill

AI Validation & Testing Chatbot & RAG Testing

• LLM/Prompt Evaluation
• Python Validation
• AI Solution Testing: Validates AI-powered solutions including chatbots and generative AI features for accuracy, reliability, and edge cases
• RAG Validation: Tests Retrieval-Augmented Generation pipelines for retrieval relevance, grounding, and hallucination handling
• Agentic AI: Builds and tests agentic solutions such as MCP and prompting

Test Data & AI Evaluation Data Quality

• AI Eval Metrics
• Responsible AI
• Test Data Management: Curates and versions golden/evaluation datasets and validates data pipelines for integrity, completeness, and drift across training and inference
• AI Evaluation: Hands-on with evaluation frameworks and metrics such as RAGAS, DeepEval, and LLM-as-judge, covering groundedness, relevance, and precision/recall
• Responsible AI & Privacy: Bias and fairness checks, safety/guardrail, and PII/data-leakage validation

Languages & Frameworks Python

• Pytest
• Core Languages: Proficient in Python for automation
• Test Frameworks: Hands-on with Pytest, or equivalent
• Framework Design: Build, scale, and maintain automation frameworks end-to-end

Secondary Skill UI & API Testing UI Automation

• REST/GraphQL APIs
• Contract Testing
• UI Automation: End-to-end browser testing across modern web applications. Well versed with BDD frameworks like Cucumber/SpecFlow
• API Testing: Validates REST/GraphQL APIs across microservices for correctness and edge cases
• Cross-layer Coverage: Bridges UI and API layers for integrated test coverage
• Tools: Experience with tools like Postman/Insomnia/Newman

Cloud, Infrastructure & CI/CD AWS

• SQL
• CI/CD Pipelines
• Cloud Exposure: Working knowledge of AWS/Azure services for test environments
• SQL Proficiency: Uses SQL to query, validate, and debug data confidently
• CI/CD Integration: Integrates test suites into Jenkins, GitHub Actions, GitLab CI, etc.