LLM QA

Key Responsibilities
Test Planning & Execution Design and execute comprehensive test plans for applications integrating LLMs and AWS services. Validate data pipelines, model outputs, APIs, and user interfaces with a focus on LLM-driven features such as text generation, summarization, and classification. LLM Quality & Evaluation Develop test cases for LLM responses, ensuring accuracy, relevance, consistency, and compliance. Evaluate prompt quality, model outputs, and fine-tuned model behavior using both manual and automated techniques. Support RAG pipelines and validate integration with vector databases or knowledge sources. Automation & Tooling Build and maintain automated test suites for backend services and LLM workflows using tools like PyTest, Postman, Selenium, or Playwright. Integrate tests into CI/CD pipelines (e.g., using GitHub Actions, Jenkins, or CodePipeline). AWS Integration Testing Test cloud-native services and workflows built with AWS tools such as: Lambda, S3, API Gateway, Step Functions, Bedrock, SageMaker, CloudWatch Validate infrastructure-as-code using CloudFormation or CDK in test environments. Performance, Security & Compliance Testing Conduct performance testing of LLM APIs (e.g., latency, throughput, model fallback behavior). Ensure compliance with data privacy, audit, and security standards when using LLMs in production workflows. Collaboration & Documentation Work closely with engineers, data scientists, and product managers to define quality metrics for LLM use cases. Document test strategies, bug reports, and QA process improvements. Technical Skills 5+ years of experience in software or data quality assurance. Hands-on experience testing applications built on AWS (Lambda, S3, API Gateway, CloudWatch, Step Functions, etc.). Experience testing or validating LLM-powered features (e.g., OpenAI, Amazon Bedrock, HuggingFace, Claude). Strong skills in test automation (e.g., PyTest, Selenium, REST API testing). Good understanding of CI/CD and DevOps in AWS environments. Experience with Python or JavaScript/TypeScript for test scripting. Nice to Have Familiarity with LLMOps, prompt testing frameworks, or prompt injection scenarios. Experience with LangChain, RAG patterns, or vector databases (e.g., Pinecone, FAISS). Exposure to model evaluation metrics (e.g., BLEU, ROUGE, factual consistency). Basic knowledge of data pipelines, ETL, or analytics workflows. Preferred Experience Testing GenAI or AI-enhanced applications in real-world environments. Ensuring ethical AI usage, data privacy, and compliance in AI workflows. Familiarity with monitoring LLM behaviour in production and logging tools.