Pandacorp
AI-operated software factory · since June 2026
A software factory I built and run on my own. I give it a product idea, and AI agents do the work under my direction. This portfolio is one of the products that came out of it.
- The problem
- Working alone, documentation and tests are the first things to slip.
- What I did
- I built a reusable process with shared rules, documentation and mandatory checks.
- The result
- A complete software factory, run by AI from idea to tested product.
Building a complete product on my own
Building products alone usually forces a trade-off: ship fast and let documentation, tests and consistency slip, or slow down to a crawl to keep them. I wanted every project I build, including this portfolio, to go from idea to a documented, tested product ready for production without a team, and without re-deciding the same engineering questions from scratch each time.
One process I can use across projects
I built a process I can reuse across projects, with shared standards, agreed engineering rules and checks that stop work when something fails. AI agents work through research, requirements, design, architecture, development, review and release under my direction.
Shared rules and lessons for the next project.
Projects share a set of engineering rules covering naming, tests, security and decisions that need my approval. Each build also records useful lessons: a library that fell short, a problem that took an hour to solve or an approach that worked. Those notes are available to the next project.
The product is documented before development starts.
I start with a rough product idea. Before coding, the factory researches the problem and similar apps, then prepares the product specification, requirements for each feature, visual design and technical plan. Work orders break that plan into tasks with clear acceptance criteria. I review those documents before development starts.
Every check must pass before work is complete.
An agent saying it is finished is only the start of verification. A separate step checks formatting, code quality, types and unit tests, looks for unused code and circular imports, and opens the pages in a real browser. A missing check counts as a failure. I added that rule after tests passed on a build with an empty page.
Three projects, run by one person
The factory has carried three ideas from a blank repo through specification and design, and two of them all the way to a finished, tested build. The site you are reading is one of them. Of everything the factory evaluates, only ideas that survive an explicit viability check get built at all.
I can take an idea through to a documented, tested product on my own. Shared rules and review points make that process repeatable.