AI/ML Systems Engineer — LLM Inference & Agent Infrastructure
I’m currently pursuing a PhD in Computer Science where my work focuses on ML Systems and Performance Optimization.
My research interests are centered around making modern AI systems faster, more efficient, and easier to scale. I’m especially interested in LLM serving, distributed inference, GPU performance, prefill/decode disaggregation, KV-cache optimization, speculative decoding, batching and scheduling, model parallelism, and high-performance AI infrastructure.
Before starting my PhD, I worked as a Machine Learning Engineer, where I became interested in what happens beyond model training: how AI models are actually served, optimized, and deployed in real-world systems.
Right now, I’m exploring the systems challenges behind large language model inference, from reducing latency and improving throughput to designing better serving architectures for the next generation of AI applications.
Inference & serving · vLLM · SGLang · TensorRT-LLM · Triton Inference Server · CUDA · paged attention · speculative decoding · KV-cache management · AWQ/GPTQ/FP8 quantization
Distributed systems · NCCL · RDMA · NVLink · FSDP · DeepSpeed · Ray · Slurm
Agents & LLM · PyTorch · Hugging Face Transformers · LangGraph · LlamaIndex · MCP (Model Context Protocol) · Anthropic API · OpenAI API · RAG
Platform & MLOps · Kubernetes · Docker · AWS (SageMaker, Bedrock, Lambda) · GCP (Vertex AI) · Prometheus · Grafana · OpenTelemetry · MLflow · GitHub Actions CI/CD
Languages · Python · C++ · Rust · CUDA · SQL · Bash
Working on inference infrastructure, agent systems, or anything at the boundary between research and production AI? I'd love to talk.