AI/ML engineer (Part Time)

Job Description

About OnSite IT

OnSite IT is an IT support & consulting company in Ocala, Florida. We support software development teams and clients around the world, and currently we require experienced AI/ML engineer for our team. The team is small and works remotely.

The role

We are hiring an experienced AI/ML engineer for less than 5 hours a week.

This is for someone who already has a full-time engineering job and wants extra income from a small, well-defined commitment. You will not be on call, you will not sit in standups, and you do not need to be free during our working hours. The work happens in writing, on your own schedule.

We have people who can build. What we are short of is someone who has already run LLM systems in production and can tell us where ours will break before we find out the hard way.

The work

Go through our AI and product roadmap with the founders every few weeks and say which items are worth doing next.

Design the architecture for our AI features and LLM systems: how the model gets called, how data is retrieved, how prompts are tested, and what happens when the model fails, slows down, or costs more than planned.

Read our code and pull requests and write back what needs to change, covering security, how the system will hold up as it grows, and whether the next engineer can maintain it.

Help us choose tools. Which model provider, which vector database, which framework, how we deploy and watch it in production. Telling us which ones we do not need is just as useful.

Join project planning when we scope new work, point out the parts that will take longer than we expect, and answer our engineers' questions in writing.

Most of this is written reviews, short architecture notes, and replies in our chat. Write code when that is faster than explaining, but code volume is not what we are hiring for.

Who we are looking for

You have built LLM systems that real users used, and you have handled what goes wrong after launch: poor output quality, slow responses, rising API bills, prompts that stop working when the model changes.

You have a strong full-stack background and have shipped and maintained production systems.

You have designed systems where the architecture choices mattered, and made tradeoffs with a small team and a limited budget.

You know your way around AWS, GCP, or Azure, along with CI/CD and normal deployment practice.

You write clearly. Nearly everything here happens in writing, so a paragraph a junior engineer can act on is worth more to us than time on a call.

You have guided other engineers before.

A degree in computer science or engineering is welcome. What you have built matters more.

Terms

Part-time, under 5 hours per week, ongoing.

You keep your current job. Please check that your contract allows outside work, and tell us about any conflict of interest before we start. We will treat your employer's confidentiality as seriously as our own.

How to apply

Tell us about one LLM system you took into production: what it did, what broke or ran badly first, and what you did about it. A short paragraph is enough. Include how many hours a week you can commit and your rate.