Model Optimization Engineer @ Roblox (2025 – Present)

Systems ML Engineer @ Meta (2022 – 2025)

• Contributing to Triton GPU compiler + PyTorch integrations with optimizations for GPU kernels (FlashAttention v2, FlexAttention) targeting H100 and B200 GPUs
• Developing pure PyTorch-based preprocessing transforms for internal customers (Ads, IG, MRS, Feed) and handling model serialization, model splitting, and export via TorchScript and FX-trace
• Authoring and optimizing PyTorch C++ operators and CUDA kernels for accelerated GPU computing

Software Engineer @ AWS Inferentia/Trainium (2021)

Building distributed training infrastructure for AWS Inferentia and Trainium. Developed Neuron SDK with graph capture and optimizations for FX-graph/TorchScript in PyTorch and TVM on TensorFlow.

Research Engineer @ AWS Rekognition (2020)

Designed and deployed anomaly detection models in PyTorch for manufacturing defects for Amazon Lookout for Vision. Built a fault visualization tool to validate model performance and feed corrections.

Machine Learning Engineer @ Citi (2017–2019)

Solved NLP and CV use cases using deep learning and traditional vision techniques. Utilized NLP for parsing financial documents and extracting relevant information.

Google Summer of Code – Mentor (2018–2019)

• TensorFlow (2019): Mentored Ryan Lee in developing a curiosity module for TF-Agents with Oscar Ramirez, implementing Random Network Distillation (RND)
• Scilab (2018): Mentored Soumitra Agrawal on a machine learning toolbox for Scilab
• Gensim (2018): Co-mentored Aneesh Joshi on neural network similarity learning for Gensim under NumFOCUS

Google Summer of Code – Student (2017)

Developed a modular ML toolbox for Scilab enabling remote execution of Python model training over the network with local inference. Project page