About

Manav C. is an Applied Science Lead at Zalando with a passion for deep reinforcement learning and a strong background in data science and software engineering. With over 5 years of professional experience, he has developed expertise in time series forecasting, causal inference, and image segmentation. Manav leads a team of 5 Applied Scientists at Zalando, where he develops and maintains the company's proprietary forecast product. He uses cutting-edge deep learning algorithms and techniques to create a multi-horizon forecast product for multiple countries, analyzing the impact of endogenous and exogenous factors. Manav also teaches a tutorial series on the theory and applications of Deep Reinforcement Learning to the Zalando Applied Science community and mentors colleagues to join the Applied Science job functions. Prior to his current role, Manav worked as a Research Intern at idsia, where he conducted research on meta-learning in a supervised setting. He also reproduced the results of 'Learning to learn with gradient descent' by Hochreiter et al., 2001, and implemented a custom LSTM cell architecture to improve the performance of the meta-learner. Manav completed his Master's degree in Artificial Intelligence from USI Università della Svizzera italiana in 2020, where he worked on model-based hierarchical reinforcement learning for his thesis. He also modified the Self-consistent Trajectory Auto-encoder (SeCTAR) to work with image-based environments and developed a deep active exploration strategy in RL using trajectory auto-encoders. Overall, Manav is a dedicated researcher and leader who is eager to learn and stay up-to-date with the latest developments in the wider forecasting community. With his expertise in deep reinforcement learning and data science, he is well-positioned to drive innovation and solve complex problems in his field.