Sankalp Mehani

Experience

• 

  Vedisoft Software and Education Services Pvt. Ltd.

  Apr 2019 - Jun 2019

  Intern

  • Developed immersive Unity-powered 3D game in C#, driving positive user feedback and substantial increase in engagement.• Led transformative web interface overhaul, enhancing user satisfaction and workflow efficiency.

• 

  Apple

  Sept 2020 - Aug 2022

  Software Engineer

  • Architected PortalPlus application for Apple TV by collaborating closely with the design team, achieving a 30% boost in user retention and a 25% increase in satisfaction.• Led development of RESTful APIs in JavaScript/Node.js, optimizing SQL with Knex.js and streamlining CRUD operations with Bookshelf.js, while integrating AWS services such as EC2 and S3, resulting in 40% reduction in server response times.• Engineered and maintained high-performance PostgreSQL databases, enabling seamless handling of 100% user growth in a year while maintaining 99.9% uptime.• Spearheaded a 6-member back-end development team, driving project-planning and execution to deliver results 20% faster.• Streamlined onboarding by mentoring 4 junior developers in software architecture, database design, and performance optimization, increasing team productivity by 50%. Show less

• 

  University at Buffalo

  Feb 2024 - now

  Research Assistant

  • Orchestrated the development of CANDO.py, an open-source platform for predictive bioanalytics, designed to analyze large-scale drug-protein-disease data and facilitate drug repurposing and discovery. • Engineered and streamlined data pipelines using numpy, pandas, and SQLAlchemy, reducing data processing time by 40% and enabling analysis of over 1 billion drug-protein interaction records.• Leveraged rdkit for molecular fingerprinting and similarity scoring, improving accuracy of drug-protein interaction predictions by 25% and aiding in identification of potential drug candidates.• Designed and implemented machine learning models, including Random Forest, SVM, and Logistic Regression, achieving a 24% increase in prediction accuracy for novel drug candidates across multiple diseases. Show less