About

Devesh Singhal is a highly accomplished software engineer with over 5 years of relevant experience in the field. Currently working as an SDE III at Google, he is responsible for developing and maintaining high-performance software systems that are used by millions of users worldwide. With a strong focus on machine learning, he has worked on optimizing bid calculations and CTR predictions in a low latency pipeline, using multiple models and statistics-driven algorithms to make better predictions than competitors in the contextual advertising space. Prior to his current role, Devesh worked as a Machine Learning Developer at Directi, where he was instrumental in increasing profits from online advertising. He worked on optimizing bid calculations and CTR predictions in a low latency pipeline, using multiple models and statistics-driven algorithms to make better predictions than competitors in the contextual advertising space. Devesh is also a skilled researcher and has worked at Indraprastha Institute of Information Technology, Delhi as a Researcher. He worked with Prof. A.V. Subramanyam to develop OneStopTuner, an end-to-end architecture that auto-tunes JVM parameters to optimize performance metrics. He successfully used active learning paradigms to reduce the amount of data required and used sample-efficient methods such as Bayesian optimization to select the best JVM flag configurations for optimizing the chosen user metric. Devesh was able to optimize execution time up to 1.35x over default Spark execution, as compared to 1.15x speedup by using flag configurations proposed by standard algorithms like simulated annealing. At the same time, he was able to reduce the number of executions for data-generation by 70%, which enabled him to suggest the optimal flag configuration 2.4x faster than the standard simulated annealing-based approach. Devesh has a Bachelor of Engineering in Information Technology from NSIT and an AISSCE from Sanskriti School. He is proficient in various programming languages such as Python, Java, and SQL, and has expertise in tech stacks such as Spark, Reinforcement Learning, AI, CNN, Deep Learning, NLP, Scikit, Data Scientist, TensorFlow, Neural Networks, Kafka, and Azure. With his vast experience and knowledge, Devesh is a valuable asset to any organization.