Domino: some pills are treating other pills

What the whole team knows

Prescriptions, findings, review tasks, notes, signed decisions and conversations live in one FalkorDB graph per clinic. That makes the clinic queryable as a team: who prescribed what, where cascades cross from one prescriber to another, who has handled each kind of cascade, what the agents learned from the team, and one search over everything written down. Every answer below is a Cypher query; the chat cites them.

827

open cascade links

61%

cross from one prescriber to another

371

downstream drugs added at walk-in clinics

tasks and decisions by people

procedures learned from the team

Where cascades cross prescribers

Each ribbon is a set of open cascade links: the first drug written by the prescriber on the left, the drug that treats its side effect added by the one on the right. Colour is the kind of prescriber who added it.

Family doctor Walk-in clinic Specialist320 links stay with one prescriber

Synthetic data: the generator assumes each patient has one family doctor and some drugs are added elsewhere.

Ask the care-team brain

gemini:gemini-3.8-flash · tools via MCP · Riverside only

Questions about the whole clinic: what the team decided and why, who has handled a kind of cascade, where prescribers hand off. Every answer cites the graph queries behind it, and your conversations are remembered.

Who knows what

People who have decided on each kind of cascade: tasks they acted on or dismissed, and decisions they signed.

Sam Lee13 decisions
ACE inhibitor → dry cough → cough suppressant · 5Calcium channel blocker → ankle swelling → diuretic · 4NSAID → high blood pressure → antihypertensive · 4
Dr. Amara Okafor4 decisions
Calcium channel blocker → ankle swelling → diuretic · 4
J. Morin, PharmD1 decision
Metformin → low vitamin B12 → B12 supplement · 1
Dr. E. Okoro1 decision
Lithium → underactive thyroid → levothyroxine · 1

Team notes

Protocols, handovers, chat messages and minutes, in the same graph as everything else. Each note is linked to the drugs, symptoms and patients it names, so it shows up in a patient’s memory, in search, and in what the agents read before a review.

Search everything the team wrote down

Review tasks, drafted notes, team notes, signed decisions, findings and your own conversations, each tied back to its patient. Found by words (FalkorDB’s full-text index), by meaning (embeddings stored on the same nodes, in FalkorDB’s vector index) and through the graph, in one ranking.

Who owns what

Every prescriber in this clinic’s graph, the patients they see, and how many cascade links start or end with a drug they wrote.