Data Analyst (Python, R) at NielsenIQ

Grade level: T4  Superior’s Position: RD GOT Innovation Leader (Business Process Architect)  As a member of the Innovation team, you will work in a dynamic, multinational, and virtual environment collaborating with colleagues across Data Ops and Global Operations Transformation teams globally.  Purpose of the job:  Data Analyst/Engineer oversees the optimal performance of machine learning (ML) models within their life cycle. Their primary responsibilities involve continuous monitoring, analysis, and feedback provision to enhance the overall efficiency of ML models. Here's a summarized breakdown of their role:  RESPONSIBILITIES:  • Continuous Monitoring: Data analysts constantly monitor the performance of ML models in real-time. This involves tracking key metrics such as accuracy, precision, recall, and other relevant indicators using existing KPI dashboards.   • Data Analysis: They analyze large volumes of raw data generated by ML models (prediction and feedback), identifying patterns, trends, and anomalies. This analysis helps in understanding the behavior of models under different conditions and detecting any potential issues. The final goal of their job is to identify a root cause that led to a drop in the data trend.   • Feedback Loop: Establishing a feedback loop is a critical aspect of their role. Monitoring the feedback of results resulted from each model on each iteration. Data analysts provide timely feedback to ML engineers and developers about the model's performance, suggesting improvements or adjustments to enhance accuracy and efficiency.   • Model Tuning and Optimization: Based on their analysis, they collaborate with the development, ML research and MLOps teams to fine-tune and optimize ML models. This may involve adjusting hyperparameters, incorporating new data, or retraining models to improve overall performance.   • Anomaly Detection: Identifying and addressing anomalies in model behavior is a key responsibility. Data analysts work to detect issues such as drift in data distribution, unexpected bias, or performance degradation, taking corrective actions when necessary.   • Collaboration with Cross-functional Teams: They work closely with various teams involved in the ML development lifecycle, including ML engineers, developers, and operations teams. Effective communication and collaboration are essential to implementing improvements seamlessly.