Tushar D.

Bioinformatics Software Engineer @NMDP

Portland, OR, US

Design and development of scalable software systems and machine learning pipelines that accelerate immunobiology research and improve transplant outcomes for diverse patient populations- Optimized a core transplant outcome prediction model, reducing runtime from weeks to hours through algorithm refactoring, parallelization, and efficient data pipeline design — enabling rapid iteration for clinical and research use- Spearheaded a Kaggle competition to advance equitable ML models for transplant survival prediction, promoting open science, fairness in AI, and reproducible research across underrepresented ancestral groups- Building a sample inventory integration application that links donor/recipient biosamples to transplant events and research cohorts, improving data traceability and supporting longitudinal genomic studies- Designed and deployed bioinformatics pipelines in Python and R, containerized with Docker and orchestrated on AWS and on-prem Kubernetes and batch processing frameworks- Contributed to a peer-reviewed publication on HLA-mismatched donor availability across ancestries, demonstrating the real-world impact of engineered systems on transplant equity.Technologies: Python, R, Java, PyTorch, LangChain, LLMs, AWS, Docker, Kubernetes, CI/CD, GitHub, GitLab, SQL, NoSQL, Batch Processing, Distributed Systems

Visvesvaraya Technological University

Bachelor of Engineering (BE), Computer Science

University of Central Florida

Masters, Computer Science