Sense. Think. Act. Learn.
Opian works at the intersection of health systems, data, and software. We build tools that structure clinical and supply chain information so it can be used at the point of decision — by the people already doing the work.
Four phases of a decision.
Health decisions — whether about drug supply or patient care — follow a pattern. Information is gathered. It's interpreted. Someone acts on it. And what happened feeds the next round. These phases are often disconnected. Our work sits in the gaps.
Sense
Structured data collection at the source — facilities, pharmacies, consultation rooms. Replacing paper tallies and disconnected spreadsheets with a shared digital record.
Stock levels across 5,100 facilities. Patient vitals at triage. Consumption figures reported quarterly.
Think
Computation over that data — forecasting models, clinical algorithms, pattern detection. Not to replace judgment, but to give it better material.
A supply model flags emerging shortages. A clinical algorithm suggests differential diagnoses based on what's been entered.
Act
Presenting the result where it's needed — to the planner adjusting a procurement order, or the clinician choosing a workup. The person decides. Always.
A planner redistributes stock before a shortage. A clinician reviews a suggested protocol and modifies it.
Learn
Recording what was decided and what happened next. Outcomes flow back into the models — so every planning cycle and every consultation sharpens the one after it.
Forecast accuracy measured against actual consumption. A protocol clinicians keep modifying gets re-examined.
We're interested in what happens when you combine health domain knowledge, locally grounded data infrastructure, and software that adapts to how people actually work.
Not a platform for everything. A specific bet: that structured, timely information — delivered inside existing workflows — compounds into better decisions over time.
In practice
Two systems so far.
ForLab+
Automates the government supply planning process that was previously done in spreadsheets. ~7,000 users across 5,100 Ethiopian health facilities. Running since 2022.
Link
Clinical workbench for Ethiopian hospitals and health centres. 15 modules covering registration through pharmacy, with a decision-support engine for 200+ primary care conditions. In testing.
What changed
Measured, not projected.
Every number describes what's deployed and working today — not a projection. Read the case studies.
Based in Addis Ababa.
Our work started in Ethiopia — a country with 5,100+ public health facilities, a national health insurance programme, and a government supply chain that plans quarterly. That context shaped everything we've built. We're now looking at where else it applies.
The questions ministries ask first
Sovereign by design
Country data stays in-country. On-premise or ministry-preferred infrastructure — no foreign cloud dependency.
Data protection
How patient and facility data is handled — and who owns it: the ministry, always.
Who we work with
Ministry of Health and EPSS in Ethiopia, funded primarily by the Gates Foundation — and diversifying.
Our position on AI
The AI story is not the model — it's the millions of interactions. Judgment stays with people.