Machine Learning Quality Control

PLEASE NOTE: there is a problem at the bottom of the posting. Please solve it and provide the answer. Please don't reply with messages like:"I can solve this problem". You won't be interviewed for sure, let alone hired. You must provide a reasonable explanation why you are a good fit AND solve this problem. Otherwise, please don't waste your time. Friendly is a robotic cognition startup specializing in document processing and workflow automation. We're using cutting-edge technologies in vision and NLP to raise fields like document analysis and information extraction to an entirely new level. Our company is built on a culture of self-organizing teams, innovation and respect. We're seeking an experienced, highly analytical individual with a data science background to join our QA team. *The ideal candidate will:* Develop automated processes and perform sophisticated testing and analysis to ensure consistent quality of final results Inspect final results and prioritize issues based on business value Build and manage Kubernetes clusters either on-premise or off-premise Collaborate with and enhance the QA team by exploring novel ways to measure accuracy Understand how machine learning models are measured, and possess a basic understanding of neural net training processes Qualifications: Design and implement unit and integrations tests covering machine learning models with a variety of data sets Create statistical robustness tests for trained models Create simulated dataset replicating real-world scenarios Debug and define corrective action Create detailed, comprehensive and well-structured test suites using programming languages such as Python Track quality assurance metrics Understanding of Data Science/Machine Learning models Ability to debug/troubleshoot large sets of data related issues and formats Creating a measure to quantify the quality of data and models. Design and Develop validation framework for ML models. Collaborate with Data Scientist to understand data and models. Applying statistical concepts to validate and QA data and models Please ALWAYS include resume and please provide the answer to the attached test. You won't receive a reply otherwise.