Apple or Android?
Most dosing methods work like Apple. Polished and simple, but closed: their products, their lab, their rules. OpenDose works like Android.
Most methods
- Their own product line
- Their approved lab
- A spreadsheet decides your doses
- Step outside it and you're on your own
OpenDose
- Any product, raw chemicals included
- Any ICP lab
- Your AI reasons with you, you decide
- Built to be changed and made your own
Religion or science?
Some methods ask for faith. Follow the rules, don't question them, trust the tool. OpenDose works like science.
Faith
- Fixed targets handed down
- Don't deviate, don't guess
- Big corrections after each test
- Results stay in a private group
Science
- Test, observe, adjust, repeat
- Every suggestion shows its reasoning and risk
- Small daily doses matched to the tank
- Results shared and challenged in the open
If you want a tried-and-true recipe, a fixed method is fine. If you want to understand your reef, start here.
How it started
Blended trace products moved several elements at once, so some crept up while others stayed too low. A fresh mix of one of the most common salts tested all over the place on ICP. The fix was to stop following and start measuring: dose what the tank actually uses, change things slowly, and let the next test decide.
A year in, the numbers got steadier every month, and every move made sense.
Why this is only possible now
Doing it this way means holding a lot in your head. Dozens of elements. Products that carry hidden extras. How kalk and two-part trade off. What changed the week before a spike.
A spreadsheet can't reason about any of that. A good AI chat can. It remembers your whole tank, explains its thinking, tells you the risk, and waits for your OK.
A few years ago, this method would have needed a chemist on call. Now anyone with an AI chat has one.
How everything else got attached
Once the AI understood the tank, the next step was obvious: let it see the dosers, monitors and lights too. That became the App Creator prompt, so anyone can build their own.
Then the bigger idea: every tank running OpenDose learns something the others could use. That's why there's a community.
What one tank can't tell you
Every tank is its own experiment: different salt, livestock, lights, food, dosing and testing. One tank can show you what happened. It can't tell you why, because too many things change at once.
The patterns that matter are quieter than that, and you only see them across hundreds of tanks recorded the same way. Questions like:
- Which elements a tank uses faster as it fills with stony corals.
- Which salts drift from batch to batch, and by how much.
- What tends to show up in the numbers before corals start to struggle.
That data has never existed in one place. Lab results sit in private databases with no context: no doses, no livestock, no outcomes. The context sits in forum threads and spreadsheets nobody can compare. Even the labs disagree on the same water.
OpenDose keeps both together, in one format, with the lab named. Every tank that shares makes the picture clearer for everyone.
Science means being honest
OpenDose isn't a proven recipe, and it carries more risk than one. One wrong decimal can hurt a tank. Change things slowly, check the math, and stay in charge.
Build it with us.