A working core first
For ADE the order was data → clustering → fuzzy score → recommendation → report. For Tideline it was app list → leftover scan → review → move to Trash. Heavy architecture is not built for an unvalidated idea.
About
We don't belong to a domain. We belong to a standard.
Sunsato builds software, systems, and tools without regard for which domain they land in. What connects the products is not a shared market but a shared engineering standard.
The ecosystem thesis
Why an ecosystem
A product standing alone has to solve everything by itself. Connected products share an engineering floor instead: the same type discipline, the same safety habits, the same release rigour.
That is why ADE, Tideline, and Vulcanus are written under the same rules despite looking at very different problems. One is decision intelligence, one is a system tool, one is a command-line product — yet all three treat the deterministic core, the honest status report, and reversibility the same way.
The Piron is the client-facing side of that same floor. Delivered projects show the structure works in practice, not only in theory.
Relationship types
Structure
The relationships under Sunsato are of different kinds; they are not flattened into one list.
- ADE
- Product — adaptive decision engine
- Tideline
- Product — native macOS application manager
- Vulcanus
- Product — command-line agent-readable memory vault generator
- The Piron
- Service brand — web services for clients
- Sunsato Labs
- Lab — experiment platform
Method
How a product gets built
The same order runs on every product: prove the core value first, grow the architecture after.
Real-world output
The measure is a working, demonstrable result. A feature counts as finished only when it holds up against real data.
Identity apart from implementation
What a product is gets defined independently of how it is written. Module boundaries follow the work, and abstraction never becomes more complex than the problem it solves.
Simple flow, strict validation
The user-facing side stays simple; the system side is never loosened. Input validation and state checks are the tightest part of the product.
Risky operations logged and guarded
If an operation matters, it is logged. Where files are deleted or hardware is involved, safety beats speed — Tideline preferring the Trash over permanent deletion is that rule in practice.
Research tied to an outcome
Research happens, particularly around ADE, but it connects to a usable result. Producing an academic archive is not the goal.
Able to stand alone
Each product is built to still make sense if it were pulled out of the ecosystem. The dependency is a consequence, not a design choice.
Operating model
How it runs
Sunsato is founder-led. Product decisions, the engineering standard, and release criteria all pass through one discipline — which is why the three products do not drift apart in tone or quality.