A Solo Founder Runs Five Products With Eight Agent "Departments" and €6 in Revenue
One developer in Portugal gave their one-person company a CEO, CFO, COO, lawyer and accountant made of agents, shipped weekly across five products for about €42 a month, and published the embarrassing revenue number alongside it
Continue exploring this workflow
The Strategy
João Pedro Silva Setas runs five small software products out of Braga, Portugal, with no employees and no funding. Rather than adopting an off-the-shelf agent platform, they gave their company an org chart: eight agents with defined roles, including CEO, CFO, COO, marketing, accountant, lawyer, CTO, and an "Improver" whose only job is to upgrade the instructions of the other seven. Each agent is defined by a markdown instruction file, and shared knowledge lives in a plain JSONL knowledge graph rather than a vector database. The agents reach real systems through MCP servers, the products themselves are built in Elixir and Phoenix and deployed to Fly.io, and the whole thing is driven through custom GitHub Copilot agents rather than a bespoke runtime. The total infrastructure bill across all five products is about €42 a month. The meta-agent is the interesting design choice. Instead of Setas rewriting prompts himself every time an agent underperforms, the Improver reviews and revises the other agents' instruction files, which turns prompt maintenance into something the system does to itself. They put the total build effort at roughly 40 hours spread over two months. And then there is the revenue: €6.09, from a single subscriber who signed up on day two. They published that number anyway, next to the 84 tweets and five articles the marketing agent shipped and the "less than one hour a week" they now spend on marketing. That honesty is the reason this is worth reading — it is a clear-eyed picture of what an agent org chart does and does not buy you.
How It Works
Treat the one-person company as an org chart and write down the roles a real company would have: CEO, CFO, COO, marketing, accounting, legal, CTO.
Define each agent as a plain markdown instruction file rather than code, so roles can be edited and versioned like documents.
Store shared company knowledge in a simple JSONL knowledge graph instead of standing up a vector database.
Connect agents to real systems through MCP servers so they can act rather than only advise.
Add a meta-agent whose only responsibility is reviewing and improving the other agents' instruction files over time.
Drive execution through custom GitHub Copilot agents rather than building a bespoke orchestration runtime.
Deploy the underlying products on Fly.io and keep the total infrastructure bill deliberately small.
Hand the marketing agent the recurring publishing and engagement work, and cap your own involvement at roughly an hour a week.
Results
Setas reports about €42 per month in total infrastructure cost across five products, roughly 40 hours of build time over two months, 84 tweets and five articles published by the agents in about two months, under one hour a week of their own marketing time, and €6.09 in revenue from one subscriber. All figures are self-reported and have not been independently verified.
Our Take
We are including this one precisely because the revenue number is €6.09. The directory is full of five-figure claims, and it is genuinely useful to see the same architecture published by someone with no incentive to flatter it. The design ideas are good: markdown role files instead of code, a flat-file knowledge graph instead of premature vector search, and above all a meta-agent that maintains the other agents so prompt upkeep is not a permanent tax on the founder. What it does not do is sell anything, which is the recurring lesson across every multi-agent company in this directory — agents are relentless at output and indifferent to whether the output reaches a buyer. Read it as an architecture reference and a warning about where agent effort quietly goes.
Frequently Asked Questions
The practical questions a builder or operator is likely to ask before trying a strategy like this.
What does this saas workflow automation AI agent actually do?
This saas workflow automation AI agent is a real workflow where the agent takes on an operational job, not just a brainstorming task. A Solo Founder Runs Five Products With Eight Agent "Departments" and €6 in Revenue shows what that looks like in practice. One developer in Portugal gave their one-person company a CEO, CFO, COO, lawyer and accountant made of agents, shipped weekly across five products for about €42 a month, and published the embarrassing revenue number alongside it The practical value comes from the agent handling repeatable business work with enough autonomy that a human only steps in after context has already been gathered.
Who should use a saas workflow automation AI agent like this?
This example is most relevant for saas operators. It is especially relevant for businesses where speed to lead, after-hours coverage, or consistent intake quality directly affects revenue. The category here is Workflow Automation, which means the best fit is a team looking to turn a manual bottleneck into a repeatable system with a saas workflow automation AI agent.
Which tools are used in this saas workflow automation AI agent setup?
The source names Copilot, Fly.io. That matters because one of the strongest signals in this directory is whether the operator shared the actual stack. Named tools make a saas workflow automation AI agent strategy far more useful than vague claims about “an AI system” doing the work.
How hard is it to implement a saas workflow automation AI agent like this?
Advanced difficulty is the current read. The listing suggests a launch window of months. Startup cost is listed as $50-200/mo. We were able to extract 8 concrete workflow steps from the source. We would treat a saas workflow automation AI agent like this as a workflow that needs real business context, testing, and exception handling rather than something you should copy blindly from one prompt.
What results can a saas workflow automation AI agent produce?
Setas reports about €42 per month in total infrastructure cost across five products, roughly 40 hours of build time over two months, 84 tweets and five articles published by the agents in about two months, under one hour a week of their own marketing time, and €6.09 in revenue from one subscriber. All figures are self-reported and have not been independently verified.
How credible is this saas workflow automation AI agent case study?
Right now the evidence comes from an article from dev.to. That is enough for us to study and curate the workflow, but not enough on its own to treat this saas workflow automation AI agent like an audited case study. We look for named tools, concrete results, and enough workflow detail to understand what was actually deployed, then we add our own editorial judgment on top.
Related Strategies
More AI agent strategies you might find useful
Pieter Levels Vibe-Coded a Flight Sim to $1M ARR in 17 Days Selling In-Game Ads
Pieter Levels built a browser multiplayer flying game almost entirely with AI co…
A Roadmap From Zero to $25K Per Month Selling Automation Services
A step by step roadmap showing how an automation agency scaled from zero to $25K…
12 n8n Workflows That Replaced $3,000 Per Month in Manual Work
Twelve n8n automations eliminated $3,000 per month in manual tasks covering lead…
How a House Cleaning Company Uses AI for Hiring, Training, Quality Control, and Retention
A Plano cleaning company built machine learning models for recruiting, AI traini…
From $2K MRR to $50K in 8 Months by Finding the Right AI Niche
A solo founder scaled an AI product from $2K to $50K MRR in 8 months after five …
From Zero to $47K in 90 Days Selling AI Automation Services
A proven system for packaging and selling AI automation to local businesses, gen…
Want more strategies like this?
Get weekly AI agent case studies in your inbox.