Test Environment Engineer (Linux / Ubuntu) – AI Startup

Job ID: 40287198

Budget: $15 – $25 USD

We do not accept autobids. We must read the entire post or your bid will be automatically rejected.

We are a new AI startup building cutting-edge systems and tooling. We move fast, build fast, and test aggressively.

We are looking for a Test Environment Engineer to help manage and troubleshoot our development environments while working alongside our internal AI coding system (Codex) and the system architect (me).

This is not a passive role. You will be actively managing the test environment while code is being generated and executed in real time.

With the combination of me (the Architect), you (the Test Engineer), and Codex (our AI coder) we can move as fast as a team of ten.

The Environment

Our stack is primarily:

Ubuntu / Linux servers

Python-based AI tooling

GPU compute nodes

Local and cloud compute environments

Rapid experimentation and testing

Codex generates a lot of code, but like any AI coder it needs oversight and guardrails.

For safety and sanity, Codex never touches our real server environment directly.
It runs in a tiny little jail it can’t escape from.

Your job is to manage the real environment and help bridge the gap between Codex’s ideas and reality.

Your Responsibilities

You will be responsible for keeping our development and test environments stable while we iterate quickly.

This includes:

Managing and maintaining Ubuntu / Linux test environments

Monitoring CPU, memory, disk, and GPU performance

Diagnosing script failures and runtime errors

Troubleshooting broken packages, dependencies, and installs

Reviewing code generated by Codex before it reaches the environment

Refreshing and resetting environments when they get messy

Helping guide Codex when it runs into issues

Stepping in manually when Codex produces something broken (which it will)

In many cases the workflow will look like:

Codex generates code

You test and observe behavior

You identify problems

You guide Codex toward a fix

If Codex fails completely — you jump in and solve it

Required Skills

You must be comfortable and proficient with:

Ubuntu / Linux administration

Linux command line tools

Shell scripting basics (bash)

Diagnosing script failures and stack traces

Package management (apt, pip, etc.)

Reading logs and troubleshooting runtime errors

Understanding system performance and resource usage

You should know how to quickly answer questions like:

Why did this script crash?

Why is this package install failing?

Why is the environment behaving differently than expected?

Is this a dependency problem, environment problem, or code problem?

Ideal Experience

Nice to have, but not required:

Python familiarity
AI / ML environments
Docker or containerized environments
GPU compute environments
DevOps style troubleshooting

To show us that you read all the way through... your bid must contain the phrase "DOMO ARIGATO MR ROBOTO" so we know you read all the way.