Overview
Prompt Testing Assistant Jobs in Nairobi, Nairobi County, Kenya at Pixel Systems
Title: Prompt Testing Assistant
Company: Pixel Systems
Location: Nairobi, Nairobi County, Kenya
We are adding a Prompt Testing Assistant to our team to help us ensure the AI components we build for our clients are reliable and accurate. This role focuses on the practical side of AI integration: checking that the scripts we write actually do what they are supposed to do when they encounter messy, real world data. You will be working directly with our developers to test, break, and improve the prompts that power internal business tools.
About Pixel Systems
Pixel Systems is a small, remote software agency. We do not build flashy consumer apps or the latest social media gadgets. We focus on the plumbing of the business world: building admin panels, CRM integrations, and database migrations. Our clients come to us when their spreadsheets have become too large to manage or when their different software platforms refuse to talk to each other. We provide a straightforward service with fixed quotes and weekly demos so our clients always know exactly where their project stands.
Our philosophy is simple: working code is better than a perfect plan that never ships. We use stable, well tested technologies like Node, React, and Postgres because we want the software we build to stay functional long after we hand over the keys. We are a senior team of developers and designers who value honesty and direct communication. As a fully remote company, we have built a culture based on trust and clear documentation rather than constant meetings or supervision.
The Role
The work of a Prompt Testing Assistant is about consistency and detail. Many of our projects now involve using large language models to help our clients automate boring tasks, like extracting information from invoices or categorizing hundreds of customer support tickets. Your job is to make sure these models behave correctly. This is a part time, contract position that is fully remote, allowing you to work from wherever you are most productive.
You will spend your time running a variety of inputs through our systems and recording the results. You will look for "edge cases" where the AI might get confused or give an incorrect answer. This is not about being creative with AI; it is about being methodical. You will help us create logs of failures, suggest small changes to the instructions we give the models, and verify that a fix for one problem does not create a new problem somewhere else. It is a fundamental part of our quality control process.
What You Bring
We are looking for someone who is organized and enjoys the process of checking work for errors. You do not need a specific degree or years of technical experience, but you do need to be comfortable working with computers and learning new software tools quickly. Because we are a remote team, your ability to communicate clearly in writing is the most important skill you can have.
- A high level of attention to detail and a move slow to move fast approach.
- The ability to follow complex instructions and checklists without skipping steps.
- Strong writing skills for documenting bugs and explaining testing results.
- Comfort with repetitive tasks that require sustained focus and observation.
- A logical mindset for figuring out why a system might be producing a specific error.
- Independence and the ability to manage your own time without a manager looking over your shoulder.
- Basic familiarity with how AI tools like ChatGPT or Claude respond to different questions.
- A professional and honest approach to reporting when things are not working as expected.
Helpful Background
While this is an entry level role, there are a few things that might make the work easier for you to pick up. None of these are requirements, but they are helpful starting points for anyone joining our team.
- Experience using spreadsheets to organize information or track data.
- Familiarity with Slack, Zoom, and other remote communication tools.
- A basic understanding of what an API is or how software generally works.
- Previous experience in a role that required quality assurance or fact checking.
- Curiosity about how different words and phrases change the output of an AI model.
- Experience working in a remote environment where communication happens mostly in writing.
Working at Pixel Systems
Working here means being part of a small, focused team that avoids drama and corporate buzzwords. We do not have layers of middle management. When you have a question or find a bug, you talk directly to the person who wrote the code. This position is paid, and we offer a rate that is competitive for the role and the market. We believe in paying people fairly for the work they do, regardless of where they are located.
Since we are 100 percent remote, you will have a lot of autonomy. We do not track your mouse movements or care exactly what time of day you do your work, as long as the testing is completed thoroughly and our deadlines are met. We value people who are straightforward about their progress and who are not afraid to say "I dont know" when they run into something new. We are here to build tools that help real people, like operations managers and warehouse teams, do their jobs more easily.
Pixel Systems is a place for people who like to solve real problems without the fluff. We are more interested in your ability to spot a mistake in a data log than in your ability to give a presentation. Most of our projects last a few months, so you will get to see a variety of different businesses and how they use technology to stay organized. If you are looking for a practical way to get involved in the software world and help build tools that actually get used, this role provides that experience in a low drama, professional environment.
