🚀 Machine Learning Engineer | Systems Builder | LLM Infra
I build high-performance AI systems that actually run in production.
From optimizing LLM inference pipelines to working deep in operating systems, I enjoy solving problems across the entire stack — from kernels to large language models.
- ⚡ Optimize LLM/VLM systems for latency, throughput, and scale
- 🏗 Build end-to-end AI pipelines (RAG, agents, training systems)
- 🔍 Design intelligent systems that replace manual workflows
- 🧩 Work across systems: OS → backend → ML → infra
- 🧠 LLM inference optimization (TensorRT-LLM, vLLM, SGLang)
- 🔗 Retrieval-Augmented Generation (RAG) systems
- 🤖 Agentic workflows using LangGraph
- ⚙️ Distributed systems & Kubernetes-based deployments
Languages
Python C++ Rust
ML / AI
PyTorch TensorRT-LLM Triton vLLM SGLang
Systems & Infra
Kubernetes Docker AWS GCP
Other
AOSP Linux Kernel RAG Pipelines LoRA Quantization
- ⚡ Improved LLM performance by 60%+ using speculative decoding & KV cache optimization
- 📄 Built a VLM-based document parsing system (98%+ accuracy)
- 🤖 Developed autonomous agents for processing real-world business workflows
- 🧪 Built RAG pipelines to generate automated integration tests from codebases
- 📱 Former Android OS engineer working on kernel, SELinux & device security
- 🐧 Linux Kernel / systems programming
- 🤖 LLM / ML infrastructure
- 🛠 Developer tools & infra-heavy projects
- 🥋 I can use nunchucks
- 🏔️ Trekked to Everest Base Camp
Build things that are not just impressive — but useful, scalable, and real.