---
title: "Discussion on \"From Research to Production: What Code Needs Before It Can Ship\" | Hashnode"
description: "Discussion on \"From Research to Production: What Code Needs Before It Can Ship\". If you've ever handed off research code to an engineering team — or received it — you probably know this moment.\nThe model works. The accuracy is great. The paper is published or the PoC is approved."
url: "https://hashnode.com/posts/from-research-to-production-what-code-needs-before-it-can-ship/6a11a32e1a3cf7bffc1e8dea"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [From Research to Production: What Code Needs Before It Can Ship](https://naphatthara.dev/from-research-to-production-what-code-needs-before-it-can-ship?utm_source=hashnode&utm_medium=feed)

If you've ever handed off research code to an engineering team — or received it — you probably know this moment. The model works. The accuracy is great. The paper is published or the PoC is approved.

May 23 · 9 min read · naphatthara.dev  
Read the article

## Written by

[naphatthara](https://hashnode.com/@Naphatthara)

@Naphatthara · Tokyo, Japan

  
[View profile](https://hashnode.com/@Naphatthara)

ML Engineer & Solution Architect · Deep tech, real impact, every phase from research to production

I build AI products from every angle — model development, cloud platform, architecture, and engineering leadership.

## Responses 1

Markdown supported · ⌘↵ to post

[Sadiqur Rahman](https://hashnode.com/@sadiqur)

[May 23](https://hashnode.com/posts/from-research-to-production-what-code-needs-before-it-can-ship/6a11a32e1a3cf7bffc1e8dea/comment/6a11cf291a3cf7bffc368b10)

The cuda:0 point hit close to home. Lost half a day on a production deploy once because GPU allocation was dynamic in our Kubernetes setup and nobody had thought to make that configurable. The research code had worked flawlessly on every lab machine for months. Such a small assumption, such a disproportionate headache.
