Your own combinations, above our suggestions
Most doctors keep a list of the combinations they actually prescribe, in a notebook or an Excel sheet. No repertory has it, and no AI was trained on it. Now you can upload yours, and it shows up right where you need it.
Ask a homoeopath with twenty years of practice what they really prescribe, and you won’t get a repertory. You will get a working list: combinations, sequences, pairings that work, the mother tincture that goes with the constitutional remedy, the biochemic salt alongside it — what has worked for that complaint, in that area, for fifteen years. It usually lives in a notebook, a Word file, or an Excel sheet someone set up years ago.
That list may be the most valuable clinical document in the clinic. And no software has been able to use it, because it isn’t in any book and no AI was trained on it.
What it does
You upload your spreadsheet. From then on, when a case you are taking matches a row you wrote, that row shows up above the ranked remedies, marked as yours.
It isn’t mixed into our suggestions, and it doesn’t quietly push the ranking. It sits above them, clearly labelled. Your own experience and a book written a century ago are two different kinds of evidence, and blending them into one list would be unfair to both.
No fixed column names needed
This is the part we worked hardest on, for a simple reason. Most import features fail at the same step: you get a template, you are told which headings to use, and you are expected to reshape years of notes to fit. Nobody does it, so the feature never gets used.
So there is no template. The app works out what each column is from what is written in it. It spots remedy names and potencies, rubric wording and free notes. Afterwards it tells you what it decided each column was, and which rows it couldn’t read.
Here is what it picks up from each row, whatever your headings say:
| What it finds | What it does |
|---|---|
| Remedy names, in your own spelling and abbreviations | Matched to our remedy list where possible. If it can’t match one, it keeps exactly what you typed and never guesses a similar remedy |
| Potency written with a remedy — 30C, 200, 1M, 6X, LM | Stored the same way a prescription stores it, so it carries straight through |
| Indications — what the row is for | Used to match the row to your cases |
| Rubrics, if you wrote any | Matched to real rubrics in the books |
| Dose and notes — how you give it | Kept word for word and shown with the row |
| Priority | Your own order, kept as you set it |
Rows it can’t read are listed with their row number and the reason, such as “no remedy recognised”. Nothing is quietly dropped. A sheet that half-imports without telling you is worse than one that refuses.
It can’t change our suggestions — and we test that
The obvious worry is that your list starts steering the repertory results. A month later, you couldn’t tell which suggestions came from Kent and which were just your own habits reflected back at you. That is how a helpful tool turns into an echo chamber.
So we didn’t touch the ranking at all, and a test proves it. It runs a case with and without a matching combination list and checks the remedies, their order and their scores come out exactly the same. It also checks that your combination really did show up, so the test can’t pass just because the list failed to appear.
Your rows are shown to you. They don’t get a vote.
Who can see your list
A list like this is your own professional knowledge, often built over a whole career, and sometimes the thing that sets your practice apart. So the settings start from that.
- Private to you by default. A list you upload is visible only to your account unless you choose otherwise.
- Sharing with your clinic is your choice, list by list. Nobody else can switch it on for you.
- Several named lists — children, skin, whatever you keep separately — instead of one big pile.
- Deleting really deletes. Removing a list wipes its stored text instead of just hiding it, and the deletion is logged. In a group practice, the clinic lead can delete a list in their own clinic, and that is logged too.
- Stored encrypted. Matching uses a protected index, so everyday case matching doesn’t need to unlock the full text of your list.
One thing we won’t overstate. Your list is encrypted when stored, but it is not end-to-end encrypted. Our servers do the matching, so they can read the list. What we promise is that nobody else can see it unless you share it. Any company claiming server-side matching is end-to-end encrypted is claiming something that can’t be true.
The limits
Each spreadsheet can have up to 2,000 rows and be up to 2 MB, and each doctor can keep up to 20 lists. Uploading a corrected sheet keeps the same list and updates it, so nothing linked to it is lost. If your list is bigger than that, write to us — we would like to see it.
Why this matters
A lot of software talk says computers will one day know more about prescribing than the prescriber. We don’t agree, and this feature is what we built instead.
A repertory holds what its authors recorded. An AI holds a rough impression of whatever text it was trained on. Neither holds what really sets your practice apart — years of what worked, for your patients, in your climate, at your prices, with your follow-up. That knowledge exists in one place only, and no computer can create it.
What software can honestly do is carry it for you without claiming it: keep it safe, notice when a new case looks like one you have already solved, show it at the right moment, and mark it clearly as yours, so it is never mistaken for a book’s opinion, or ours.
The final decision is the same as always: yours. The difference is that you are no longer the only place your best thinking is stored.