---
title: "Harshul Jain"
category: "people"
role: "Senior Software Engineer - ML/AI"
company: "Audible"
linkedin: "https://www.linkedin.com/in/hjain1393/"
twitter: "https://x.com/hj1393"
website: "https://harshuljain.substack.com/"
blog: "https://harshuljain.substack.com/"
sourceLabels: ["Official speaker roster", "Official conference schedule"]
---
# Harshul Jain

## Profile
Senior Software Engineer - ML/AI at [[audible|Audible]].

- [LinkedIn](https://www.linkedin.com/in/hjain1393/)
- [X / Twitter](https://x.com/hj1393)
- [Website](https://harshuljain.substack.com/)
- [Blog](https://harshuljain.substack.com/)

## Biography
Harshul Jain is a Senior Software Engineer at [[audible|Audible]] (Amazon) who builds ML and LLM infrastructure at scale — AI Search serving 10M users, a feature store processing 100K transactions per second, and LLM serving and evaluation systems powering GenAI in production. He is writing LLM Inference at Scale, a benchmark-driven handbook on GPU memory engineering, attention optimization, and production LLM serving backed by a companion repository that gained 100+ clones in its first week with zero promotion.

## Conference Sessions
- [[2026-06-29-harshul-jain-2-hr-deep-dive-on-llm-inference-at-scale-part-1-of-2]] — 2 hr deep dive on LLM Inference at Scale — Part 1 of 2 (2026-06-29, 12:10pm-1:10pm)
- [[2026-06-29-harshul-jain-2-hr-deep-dive-on-llm-inference-at-scale-part-2-of-2]] — 2 hr deep dive on LLM Inference at Scale — Part 2 of 2 (2026-06-29, 1:15pm-2:15pm)

## Evidence Graph
This evidence graph summarizes how this person appears across the conference source graph: scheduled sessions, linked videos, transcripts, and slide-derived evidence.

### Linked Sessions
- [[2026-06-29-harshul-jain-2-hr-deep-dive-on-llm-inference-at-scale-part-1-of-2|2 hr deep dive on LLM Inference at Scale — Part 1 of 2]]
- [[2026-06-29-harshul-jain-2-hr-deep-dive-on-llm-inference-at-scale-part-2-of-2|2 hr deep dive on LLM Inference at Scale — Part 2 of 2]]

### Media Signals
No linked video, transcript, or slide source has been attached yet.