Hello everyone, I am Saurabh!
Check out my blog posts about AI/Machine Learning ! Blog
I am an AI Researcher and Engineer with a Master’s degree in Machine Learning, combining 4+ years of US industry experience and 3+ years overseas in developing and deploying AI across diverse domains. Published research at leading venues including ICLR, ICCV, MICCAI, and Discovery Science in NLP, Computer Vision related to efficient AI, representation learning, robust and fair AI. I have done open source contributions through Google Summer of Code.
* Equal Contribution
- D. Moukheiber*, Saurabh Mahindre*, Mingchen Gao. Diagnosing the Effects of Pre-training Data on Fine-tuning and Subgroup Robustness for Occupational NER in Clinical Notes. Workshop on Spurious Correlation and Shortcut Learning, ICLR 2025.
- D. Moukheiber*, Saurabh Mahindre*, et al. Looking Beyond What You See: An Empirical Analysis on Subgroup Intersectional Fairness for Multi-label Chest X-ray Classification Using Social Determinants of Racial Health Inequities. Computer Vision for Automated Medical Diagnosis Workshop, ICCV 2023.
- D. Moukheiber*, Saurabh Mahindre*, et al. Few-Shot Learning Geometric Ensemble for Multi-label Classification of Chest X-Rays. Data Augmentation, Labelling, and Imperfections Workshop, MICCAI 2022.
- Haimonti Dutta, Saurabh Mahindre, Nitin Nataraj. Consensus Based Vertically Partitioned Multi-layer Perceptrons for Edge Computing. 24th International Conference, Discovery Science 2021.
- KP Seastedt, D Moukheiber, S Mahindre, Leo A Celi. A scoping review of artificial intelligence applications in thoracic surgery. European Journal of Cardio-Thoracic Surgery.
2024: Working on AI @ Oracle Netsuite AI.
I developed AI systems for intelligent matching, ranking, and recommendation tasks by building NLP pipelines that integrated LLMs, word/document embeddings, and graph algorithms, leading to over 5% improvement in r-precision for applications like candidate-job matching. I also implemented domain-specific NER models using fine-tuned LLMs, leading supervised data curation, evaluation across 100+ customer datasets, and monitoring performance metrics such as Precision@K, NDCG, WEAT scores, and robustness under distribution shifts. Additionally, I built an LLM evaluation framework for production systems, supporting toxicity and bias detection as well as traditional NLP metrics like ROUGE, METEOR, and BLEU, across text-enhancement use cases powered by models like CommandR, ChatGPT, and LLaMA.
2022: Working on Machine Learning / NLP problems at Oracle in AI Apps team.
2021: I graduated with a Masters degree in Data Science and ML from University at Buffalo.
In Summer 2020, I worked at IBM T.J Watson Research Center as a part of IBM Research's Internship programme.
I was part of the AI Compute Group where I implemented novel and compact deep neural network algorithms and demonstrated improvements using purely Boolean Functions and added optimization schemes. We tested the performance increase on various datasets and probed implementations on embedded devices like FPGAs.
Previously I have applied quantitative techniques like machine learning while working at leading payments firm Paytm till 2019.
I contribute to open-source and have done so while being supported by Google during two Google Summer of Code projects in 2014 and 2016.
I am also currently Top 2% in Kaggle ML Competitions.
I spent a wonderful time while completing my Bachelor's in Electronics Engineering at BITS Pilani Goa Campus