Teerth Sharma
// About
Engineering at the Edge of Possibility
I specialize in hardware-aware AI optimization and sparse attention mechanisms. My work focuses on the intersection of systems programming and machine learning—building infrastructure that makes neural networks run faster, use less memory, and deploy anywhere.
Hardware-Aware Optimization
Designing algorithms that understand and exploit GPU architecture.
Sparse Attention Mechanisms
Reducing computational complexity without sacrificing accuracy.
Neural Network Inference
Optimizing TensorRT-LLM and custom CUDA kernels.
Technical Arsenal
- Rust (High Performance)
- C++ (CUDA/Systems)
- Python (PyTorch/MLX)
- Assembly
- TypeScript
- NVIDIA Triton
- TensorRT-LLM
- Apple Metal (MLX)
- WebGPU
- WASM
- DirectStorage
- Docker
- Kernel Optimization
- Sparse Attention
- DSP
- Latency Optimization
- Psychoacoustics
target_orgs: NVIDIA (TensorRT) • Google (XLA/JAX) • Meta (FAIR) • Microsoft (DirectStorage)
// Mission Log
Trajectory Timeline
Chief Technology Officer (CTO) & Co-Founder
KarmicSphere Media
Jaipur, India
Leading technical strategy and infrastructure. Overseeing deployment of Linux-based systems and optimizing administrative workflows.
Bachelor of Technology, Computer Science
Manipal University Jaipur
Jaipur, India
Focus Areas: LLM Fine-tuning, Business Analysis, High-Energy Physics applications in Computing.
Writer (Self-Employed)
Remote
Remote
Published writer specializing in creative and web content, developing a strong foundation in storytelling and communication.
System
Architectures
System Manifest
Access the full technical specifications and operational history. This document contains classified data regarding mission parameters and system architecture.
// Contact & Signal
Establish Connection
> Available protocols detected: [1-4]
© 2026 Teerth Sharma. Built with passion and love by Seal Team🦭