Useful output
Can the work meet the brief, preserve important details and be easy for someone else to use?
Experiment in progress
Can AI turn a real brief
into useful work?
A Nexa Lab experiment in finding, scoping and delivering freelance work with AI doing most of the production—and people setting direction, reviewing quality and handling decisions that need human responsibility.
The idea
What happens when AI is treated as an operator, not just a writing assistant? We test whether it can understand a brief, choose a workable approach, produce the deliverable and check its own work under real constraints.
The question is not how many tasks AI can generate. It is whether the result solves the right problem, meets the agreed scope and can be delivered responsibly.
What we want to learn
Can the work meet the brief, preserve important details and be easy for someone else to use?
How much active production, checking and revision does a clearly scoped job actually require?
Will customers choose this way of working, and can the value delivered justify the total effort and cost?
Sample work
Self-initiated demonstrations of execution and presentation. These are not commissioned client projects or evidence of paid work.
A one-page studio presentation with services, work examples, a clear enquiry path and responsive layouts. The brand and project content are demonstration material.
Open the live demo ↗A four-page Vietnamese–English demonstration for industrial supplies, with product information and a structured enquiry journey. It demonstrates information architecture, not a live supplier business.
Open the live demo ↗Demo forms prepare or copy an enquiry note; they do not send email, process payments or connect to a production backend.
How the work is done
Agree what is needed, the input files, delivery format, deadline and what is outside scope.
Check the tools and execution limits. Use a relevant sample when it helps, rather than promising capability that has not been established.
AI supports production. Checks focus on the actual brief: data accuracy, visual consistency, functional behaviour or language quality, depending on the job.
Package the agreed files with clear notes, handle the agreed revisions and record what improved the result.
Current scope is assessed job by job. Samples show specific work, not a guarantee for every request. Customer materials and private work are not published as portfolio examples without permission.
The experiment continues
We are testing this model through freelance platforms and practical briefs. Publishing samples and sending proposals are steps in the experiment—not proof of customer satisfaction or income.
For a relevant question or brief, contact Nexa Lab and mention AI Freelancer.
contact@nexalabhq.com