Aryan — ML infrastructure engineer
$ whoami
ML infrastructure engineer. I build and run the serving layer for large language models — inference throughput, request routing, cost attribution, and the unglamorous plumbing between them.
- loc
- Indianapolis, IN
- edu
- MEng Electronic & Information Engineering, Imperial College London
- lang
- en / fa
$ cat about.txt
Most of my work is taking something that runs on one GPU in a notebook and making it survive contact with real traffic, real tenants, and real budgets. Batching, routing, backpressure, failure modes — the parts nobody demos.
I have a strong bias toward reading the source over reading the docs, and toward systems that fail loudly rather than degrade quietly.
$ git log --oneline work/
ML infrastructure engineer — Certara
2024—present · LLM serving platform for pharma and federal clients
Software engineering intern — Huawei R&D UK
C/C++ · secure element operating system, performance work
MEng Electronic & Information Engineering — Imperial College London
Systems, signal processing, machine learning
$ ls -la projects/
$ tree stack/
stack/
├── serving/
│ ├── vllm
│ └── litellm
├── infra/
│ ├── kubernetes
│ ├── aws-bedrock
│ ├── rabbitmq
│ └── keydb
├── languages/
│ ├── python
│ └── c/c++
└── habits/
├── reading-the-source
└── profiling-before-guessing
$ cat now.txt
Reading DDIA and Pope et al. on inference scaling. Picking apart the vLLM scheduler. Lifting five days a week. Bank fishing the White River. Getting Farsi past the point where I can only talk about food.
$ ./contact
type help — tab completes, ↑ recalls