I am on the job-market for full-time opportunities! Reach out to me at joshiketakir@gmail.com! I have earned my Masters in Computer Science (Thesis) and M.Phil with Computer Science(Thesis) from Yale.
Previously, I have worked full-time and interned at Nvidia working on compilers and virtual memory management for tiered memory systems. I have also been a researcher at Yale where my research has been on improving the efficacy of memory management in tiered memory systems. Specifically, I have worked on improving prefetching and eviction in CPU-GPU memory systems. I have proposed the use of practical machine learning solutions to build adaptable, accurate, and resource-efficient memory management solutions. While developing these solutions, I am making use of cognitive principles to solve fundamental issues in online machine learning such as catastrophic forgetting.
I received my undergraduate degree in Computer Engineering from the University of Pune. I completed my undergraduate thesis research in compiler optimizations at the Indian Institute of Technology, Bombay, where I was advised by Prof. Uday Khedkar.
News:
- I am looking for exciting full-time opportunities! Reach out to me at joshiketakir@gmail.com!
- September 2023: Serving as the Communications Chair on the board of the Society for Women Engineers(SWE) at Yale.
- May 2023: Our work on “Mitigating Catastrophic Forgetting using principles of Context-Dependent Learning” is out on Arxiv.
- February 2023: Our work on “Prefetching Using Principles of Hippocampal-Neocortical Interaction” is selected for HOTOS ‘23.
- June 2022: Received ISCA 2022 travel grant and attended ISCA in person.
- June 2022: Joined Nvidia Unified Virtual Memory Team as an Intern. Developed a new access-aware eviction policy that suits traditional and emerging applications. It achieved two orders of magnitude performance improvement over the existing policy.