AfterQuery Experts

September 2026 - Present

Remote

AI Software Engineer (Contract)

• Benchmarked Muon optimizer kernels on NVIDIA H100 GPUs, achieving 3.85-4.17x speedup and 74-76% lower latency while validating BF16 Newton-Schulz accuracy against correctness oracles.
• Created PyTorch, CUDA, and Triton MoE dispatch/combine tests covering top-k routing, expert capacity, token packing, and FP32 accumulation, with reference-output checks and calibrated speedup scoring.
• Stress-tested multi-query attention during decoding, identifying causal masking and ragged KV-cache memory-safety edge cases with compute-sanitizer checks.
• Authored and reviewed AI coding-agent tasks for Projects Corgi and Gold, assessing algorithmic efficiency, API compatibility, concurrency safety, and regression coverage; ranked among Project Corgi's top 40 contributors.

Fiserv

July 2023 - August 2024

Chennai, India

Technology Analyst - SDE

• Developed new app features and implemented Disaster Recovery using React.js, GraphQL, AWS, and GIT.
• Addressed security findings, collaborated with the AppSec team, and utilized BurpSuite for data tampering prevention.
• Gained expertise in AWS AppSync, Cognito, and authentication libraries, enhancing application security.
• Developed a security mechanism to protect the application against data tampering.
• Performed production deployments, ensuring smooth and secure application operation.

Braynix AI

September 2021 - December 2021

Remote, India

Deep Learning Intern

• Worked on medical imaging projects focused on abnormality detection and image segmentation.
• Conducted research and implemented findings from multiple sources and research papers using TensorFlow.
• Designed and developed deep learning algorithms for medical imaging applications.
• Implemented UNET and UNET with EfficientNetB4 encoder for the segmentation of X-ray images.
• Applied object detection algorithms like SSD, YOLOv5, and RetinaNet for detecting abnormalities, and used DCGANs to generate additional training data.