Mar 2023 – Jun 2024 · 1 yr 3 mos · On-site

• Building a Payments app for Saudi Arabia where I lead the Core banking team.
• Designed and built the wallet and accounting system
• Designed and built the reconciliation and settlement engine
• Designed and built a solid and robust transaction system to ensure every SAR is accounted for.

May 2022 – Jan 2023 · 8 mos

Jan 2021 – May 2022 · 1 yr 4 mos · India

• Founded a Data Science Platform for cloud-based Data Science/AI workloads, integrating Distributed Computing, Data Pipelines, and MLOps.
• Developed a unified platform to help Data Teams scale and deploy their projects efficiently.
• Achieved seamless integration of various data processes, enhancing productivity and collaboration within the team.

Jul 2019 – Dec 2020 · 1 yr 5 mos · Austin, Texas, United States

• Consulted Facebook AI Research team on Pytorch project (Implemented first draft of torch function)
• Built Nebari (Jupyterhub, Dask, Prefect, GPUs on Kubernetes cluster)
• Contributed to UNumpy
• Contributed to Dask project

Apr 2018 – Jun 2019 · 1 yr 2 mos · Hyderabad Area, India

• Built Genomics Data Science Infrastructure for major pharmaceutical companies.
• Led a team of engineers to build data analysis pipeline of genomics data.
• Set up research collaboration with IIT Madras.
• Worked on ADAM tools to process genomics data from Whole Genome Sequencing and Whole Exome sequencing.
• Improved Faster-LMM-D/GEMMA for running on clusters.

Oct 2017 – Mar 2018 · 5 mos · Tokyo, Japan

• Winner of Ruby Grant 2017.
• Developed RbCUDA project.
• https://www.ruby.or.jp/en/news/20171206
• https://www.ruby.or.jp/en/news/20180501

Sep 2017 – Apr 2018 · 7 mos

• Genenetwork Project:
• Building Genome Wise Association(GWAS) tool called Faster-LMM-D for HPC devices.
• Implemented Linear Mixed Models for GWAS analysis.
• Added multi-core support
• Faster-LMM-D is the only GWAS tool with GPU support.

May 2017 – Aug 2017 · 3 mos

• Project: Creating the fastest math libraries for Ruby by using the GPU through CUDA, OpenCL and ArrayFire.
• Built ArrayFire gem from scratch.
• ArrayFire gem is General Purpose GPU computing library that can be used for high performance computing in Ruby be it statistical analysis of big data, image processing, linear algebra, machine learning.
• ArrayFire gem has an outstanding performance considering other existing Ruby libraries that run on CPU. ArrayFire gem can also be run on clusters and handle real world problems by crunching huge datasets. The ease of using ArrayFire gem makes Ruby a viable choice for high performance scientific computing.

May 2016 – Aug 2016 · 3 mos

• Project: Port NMatrix to JRuby.
• Built Java backend of NMatrix using Apache Commons Maths, from scratch.
• Implemented NMatrix for dense matrices with double and object ( ruby objects ) data type.
• Ported mixed-models gem to JRuby which heavily uses NMatrix at its core.