# Aghilan Nathan

> Engineer @ Baseten

Location: New York, New York, United States
Profile: https://flows.cv/aghilan

I build systems that help people train large models efficiently. I talk to users constantly, shape product direction, and care deeply about abstractions that make complex ML feel effortless - from reliability to SDK design to distributed workflow simplicity.

1) ML Platform

I build and scale ML training platforms end-to-end. That means handling real customer workloads, improving reliability, and turning messy distributed systems into clean developer experiences. I’ve worked across Kubernetes, storage, networking, observability, and backend services to make large-scale training infrastructure feel boring - in the best way.

2) ML Systems

I’m not an ML Engineer/Researcher: I don’t tune models or stare at loss curves. Instead, I handle the non-ML research problems that make or break large-scale training: distributed orchestration, scaling efficiency, failure recovery, performance bottlenecks, and workflow design.

My interest is strong scaling: increase chip count and achieve near-linear throughput. I work heavily with PyTorch Distributed, TorchTitan, and modern LLM training stacks to ensure researchers can scale confidently and focus on the science - while I make the systems actually work.

3) GPU Kernels

I stay close to GPU kernels to understand performance at the metal. I use cuTile to write high-performance kernels quickly - especially where frameworks leave performance on the table (e.g., MoE, fused ops). Many of my kernels have been merged into upstream NVIDIA libraries. I optimize for leverage: meaningful gains without spending months chasing theoretical peak.

Everything ties back to the same goal: enabling the future of deep learning by giving people the tools to train efficiently - today and long term.

Contact: nathanaghilan[@]gmail.com

## Work Experience
### Software Engineer @ Baseten
Jan 2025 – Present | New York City Metropolitan Area
🤖 Training

### Software Engineer @ Wealthsimple
Jan 2025 – Jan 2025 | Toronto, Ontario, Canada
⚖️ Margin Trading

### Research @ The University of British Columbia
Jan 2024 – Jan 2024 | Vancouver, British Columbia, Canada
🎵 Multimodal ML

### Software Engineer @ LinkedIn
Jan 2024 – Jan 2024 | Sunnyvale, California, United States
🦧 ML Training Infra

### Software Engineer @ Robinhood
Jan 2023 – Jan 2023 | New York City Metropolitan Area
📈 Equities Trading

### Software Engineer @ Intuit
Jan 2023 – Jan 2023 | Edmonton, Alberta, Canada
🧾 Quickbooks Online


## Education
### Bachelor of Commerce - BCom in Computer Science and Business
The University of British Columbia

### High School Diploma
Nelson Mandela High School


## Contact & Social
- LinkedIn: https://linkedin.com/in/aghilan-nathan-3b65bb211
- Portfolio: https://aghilan.me/

---
Source: https://flows.cv/aghilan
JSON Resume: https://flows.cv/aghilan/resume.json
Last updated: 2026-04-13