A medicine and medical-supply stock system for a multi-site clinic group, covering live stock across locations, batch-level expiry control, and purchasing.
Medistock is the system a clinic group runs its pharmacy stock on. Every medicine arrives as a batch with its own expiry date and its own storage rules, and the same item sits in three different places at once, so "how much do we have" is never a single number. The product answers that question per location and per batch, warns before stock expires rather than after, and turns the answer into a purchase order.
The clinics ran medicines on spreadsheets. Stock counts per location never reconciled, batch numbers and expiry dates were kept by hand, and a reorder was usually discovered once an item had already run out. Two of those failures are expensive in a very specific way: an expired batch is written off entirely, and a stock-out on something like insulin is a clinical problem, not an inventory one.
We designed the system around first expiry, first out, and made expiry a state the interface is always showing rather than a date somebody has to go and check. Stock is modelled per batch and per location rather than as one number per item, which is what makes the expiry watchlist, the transfer between sites, and the controlled-drug register all possible from the same records. Compliance rules live on the category, not the item, so the pharmacist configures antibiotics once instead of 318 times.
Before drawing anything we went through the actual files, including the workarounds in them. The columns people had added by hand are the requirements: that is where batch tracking, cold-chain flags, and the two-person goods-in check came from.
The data model was the real design decision. Once stock is a set of batches in places rather than a quantity on an item, FEFO ordering, expiry alerts, inter-site transfer, and per-batch quarantine all fall out of one structure instead of being four features.
This is a tool people sit in for hours, so density comes before decoration and colour is reserved for status. Every screen names its next action: the dashboard points at the critical lines, the item page proposes the purchase order, reports give a write-off split by cause you can actually act on.
Four thousand SKUs across three sites is where a table stops being a table. We built the interface against that volume from the start, so filtering, batch expansion, and export behave the same in the pharmacy as they do in a demo.
Tell us what you are building, or what you already built. We will tell you honestly how we would take it to production, and what it costs. Projects start at $200, dedicated engineers at $800, and the production review is free.
