Skip to main content
MKMKConsultantGroup
Back to all essays
Healthcare1 min read

Pharma FEFO at national scale

Near-expiry redistribution, cold chain, and GS1 verification across 400+ facilities. AI that moves medicine, not slides.

Expiry is part of the transfer decision. Identify the batch; Check stock conditions; Assess receiving need; Propose a feasible transfer; Authorize the movement; Verify the receipt.
Open full-size workflow diagram
Read the workflow steps
  1. Identify the batch. Check item, quantity and expiry. Retain the source location.
  2. Check stock conditions. Inspect storage and quality status. Expose temperature exceptions.
  3. Assess receiving need. Compare demand and available stock. Allow time for intended use.
  4. Propose a feasible transfer. Check transport and handling needs. Retain batch and quantity details.
  5. Authorize the movement. Assign the responsible operator. Review unresolved conditions.
  6. Verify the receipt. Record received batch and quantity. Investigate discrepancies.
  7. Quality or handling exception: hold.
  8. First expiry, first out does not override storage or quality requirements.
Cover graphic for Pharma FEFO at national scale, Healthcare

Drug waste is a logistics failure dressed as a clinical problem. Stock expires in one warehouse while another facility runs short. Cold chain breaks silently. Price benchmarks stay locked in a binder nobody updates.

For Guyana's Ministry of Health we shipped FEFO tracking, AI-driven redistribution of near-expiry stock, cold chain monitoring, and WHO/PAHO price benchmarking, across a network that includes 400+ facilities when you count the full distribution graph.

Interoperability or nothing

Interoperability across EHR, warehouses, and pathology isn't a nice-to-have. Without it, every "AI insight" is stranded. The model can recommend a transfer, but if the warehouse system cannot execute and verify GS1 identifiers, you have generated a suggestion, not an outcome.

When the network spans hundreds of facilities, small percentage improvements in waste are national outcomes.

Near-expiry redistribution only works when freshness, location, and demand are in the same decision loop. That loop is what we build. The dashboard is how humans supervise it.

Healthcare AI that cannot move inventory is presentation software. Ship the movement.

Putting this into practice?

Explore engineering services →Our research approach →

Continue reading

Fictional threshold comparison for 1,000 shipment messages containing 100 urgent cases. Thresholds 0.90, 0.70 and 0.40 flag 40, 100 and 220 cases, missing 64, 30 and 10 urgent cases. At four minutes per review, only the 160-minute queue fits the illustrative 180-minute capacity.

AI / ML

AI confidence thresholds: size the review queue

Fictional shipment event trace: version 18 arrives, version 17 arrives late, then event 18 repeats. A durable inbox retains the late event and deduplicates the repeat while the projection stays at version 18. A later agent proposal based on version 19 is held when current state reaches version 20.

System Design

AI shipment webhooks: handle late and duplicate events

Illustrative prompt-cache comparison with a 6,000-token stable prefix and 1,000 fresh tokens per request. Four requests ten minutes apart create four five-minute writes or one one-hour write and three reads; example input costs are USD 0.0680 and USD 0.0338.

System Design

Claude prompt caching: measure reuse before choosing a TTL