Pranjul Yadav

I had the pleasure of attending a fantastic session with Pranjul. His insights into ML solution design were incredibly valuable. He shared practical tips, highlighted critical areas to focus on, and encouraged me to think deeply about edge cases, broadening my perspective and introducing me to new ML topics. Pranjul’s expertise and clear communication made the session highly engaging. I truly appreciated his guidance and am excited to apply what I’ve learned. Thank you, Pranjul.

Pranjul's mentorship provided invaluable insights into the intricacies of the Ads ecosystem. I gained a deep understanding of the end-to-end machine learning pipeline, from data preparation to model deployment and evaluation. Key takeaways included:** Data-centric approach: We explored the critical role of high-quality data, including the importance of data collection, cleaning, and feature engineering. We discussed techniques like balancing class imbalances, handling biases, and leveraging diverse data sources. Feature Engineering: We delved into various feature representation techniques, including image, text, and video embeddings, along with the importance of tabular features and their impact on model performance. Model Selection and Evaluation: We explored a wide range of models, from traditional methods like Linear Regression and Decision Trees to deep learning architectures like RNNs, LSTMs, and Transformers. We also discussed key evaluation metrics like log-loss and the importance of robust offline and online evaluation. The ML lifecycle: We examined the complete ML workflow, from data preparation and model training to deployment, A/B testing, and continuous monitoring and improvement. We discussed the critical role of feedback loops in optimizing model performance and adapting to changing user behavior. Pranjul's ability to explain complex concepts clearly and concisely, coupled with his focus on real-world applications, made the learning experience both engaging and insightful. I highly recommend Pranjul to anyone seeking to gain a deeper understanding of the Ads ecosystem and the role of machine learning in driving business success