Recro.io - Quality Assurance Engineer - API Testing (4-9 yrs)

Responsibilities :

- Develop and execute test plans and cases for AI/ML models, algorithms, and applications.

- Validate model outputs for accuracy, fairness, and reliability across use cases.

- Test AI system integration with enterprise platforms and applications.

- Perform regression testing, Performance testing, stress testing, and scenario-based validation for models.

- Ensure models adhere to performance metrics (e. g., precision, recall, F1-score).

- Monitor and evaluate model drift and retraining requirements post-deployment.

- Design and implement automated testing frameworks for AI systems.

- Leverage testing tools such as PyTest, Selenium, or specialized AI testing platforms.

- Verify the integrity, accuracy, and consistency of training and testing datasets.

- Implement data validation frameworks to identify and rectify data quality issues.

- Audit AI systems for compliance with ethical guidelines and regulatory standards.

- Ensure adherence to fairness, bias mitigation, and explainability standards.

- Maintain detailed documentation of test cases, results, and issue resolutions.

- Report QA metrics, such as defect density, test coverage, system performance, and defect escape rate.

Requirements :

- Experience with QA methodologies and tools such as JUnit, Postman, and PyTest, and proficiency in

automation testing and CI/CD pipelines for AI systems.

- Knowledge of performance testing, model evaluation metrics, and experience with AI/ML models, data pipelines, and ML frameworks like TensorFlow, PyTorch, and Scikit-learn.

- Understanding of database technologies, including MongoDB, and big data tools such as Spark and Hadoop, with basic exposure to REST API testing.

- The ability to diagnose and resolve performance issues, with programming skills in Java, JavaScript, or similar languages is beneficial.

- Familiarity with ethical AI practices and regulatory compliance standards like GDPR and CCPA, and experience with cloud platforms such as GCP or AWS is preferred.