The problem
Before ForLab+, Ethiopia's national pharmaceutical supply planning process was fragmented and time-intensive. The Ministry of Health and Ethiopian Pharmaceutical Supply Service (EPSS) relied on facility managers to submit quarterly forecasts via email, often in custom spreadsheet formats. These submissions were manually reconciled in Addis Ababa—a process that consumed 3-4 months each quarter just to consolidate and validate data across 5,100 facilities. During this reconciliation period, planners were working with data that was already a month old by the time validation began. Inconsistencies went undetected until late in the cycle, forcing last-minute corrections. Facilities couldn't see whether their submissions were accepted or flagged. The result: supply plans were frequently incomplete, inaccurate, or missed deadline, leaving procurement teams uncertain about true demand. This manual cycle also meant that EPSS had no visibility into emerging supply variance or demand signals until the next formal quarterly round. A regional pharmacy manager noticed a spike in TB medication consumption in August, but the system wouldn't reflect this insight in planning until the October cycle—two months later.
What we built
ForLab+ automated the entire forecasting workflow. We built a unified platform where 5,100+ facility managers submit quarterly demand forecasts directly into a single system, with validation rules applied at the point of entry. Each facility sees their submission status in real time: accepted, flagged for review, or pending corrections. The platform integrates consumption data from facilities' existing pharmacy management systems where available. Pharmacists see their historical consumption patterns visualized, so they can reason about their forecasts with confidence. Regional coordinators access a dashboard showing all submissions across their jurisdiction—color-coded by completeness and variance status. The core of ForLab+ is a practical intelligence layer that flags submissions outside expected ranges without requiring human judgment at scale. When a facility's forecast is 30% higher than historical consumption with no explanation, the system surfaces this. The regional team can ask one targeted question instead of reviewing thousands of submissions manually. We deployed this on Ethiopian-controlled infrastructure (EPSS's own servers), ensuring data never leaves the country. The platform runs offline-capable, so facilities in low-connectivity areas can submit when they're ready.
What changed
National supply planning cycle compressed from 3-4 months to 6-8 weeks (Q1 2026). The first formal comparison came in Q2 2025, the second full quarterly cycle using ForLab+. By Q3 2025, we measured: - Submission completeness rose from ~70% (Q1 2024, pre-ForLab+) to 94%+ (Q1 2026). - Stockout-risk forecasting error fell within ±7.1% accuracy (measured against actual dispensing in the following quarter). - Time spent by regional pharmacy coordinators on reconciliation dropped from ~2 weeks per cycle to 2-3 days. - EPSS can now flag consumption anomalies within 48 hours of detection, enabling mid-cycle corrections. Facility managers report that the real-time validation saved them time. Before ForLab+, a facility might spend 4-5 hours filling out a spreadsheet with no feedback. Now they spend 2-3 hours, get immediate confirmation their data is valid, and can move on. Regional pharmacy supervisors describe a shift in their work. They're no longer data processors—they're now decision-makers. One regional pharmacy head in Oromia told us: "We used to look at the spreadsheets EPSS sent and check them in Excel. Now I see my data live, I see what facilities submitted, and I can ask the real questions—why is this facility's consumption pattern unusual?"
""We don't spend our weeks reconciling spreadsheets anymore. We can see the numbers live and answer the question that actually matters: why did consumption change?""