* Developed Foundation Models for drug imaging data that has 5x improvement over off-the-shelf models* AI features (embeddings) for Drug Similarity and Toxicity screens used by several big pharma* Identify prototype single cells for drug phenotypes and MoAs * Developed and published methods for removing confounders in medical imaging* Designed and implemented various similarity features for Spring Science App used by biologists and data scientists* Developed efficient backend for embedding generation for billions of cells to enable biologists to use SOTA AI* Worked directly with customers to understand their pain points and implement novel technical solutions* Implemented POCs for virtual staining, multimodal feature fusion, 3D segmentation under a week

Manage and guide a team of engineers to explore novel applications of deep learning in digital pathology.

Explore novel methods in Machine Learning (architectures, loss functions, etc) for various tasks in digital pathology. Solve problems in Multiple Instance Learning, Semantic Segmentation, cell classification and other associated tasks.

Design and develop deep learning-based models for region and cell-based prediction of various end-points.

Research in Deep learning techniques in NLP, especially paraphrasing, text simplifciation and question answering.

Research application focussed in applying deep learning to problems of NLP. Designing new techniques for memory networks.

Efficient training for convolutional neural networks.

– Explored model parallelism for convolutional neural networks– Architecture learning for reduced model complexity

I worked on very large datasets and applied various distributed algorithms and machine learning techniques to derive insights.