Three practices, twelve chairs: 75% of patients now finish their treatment plan, up from 50%.

The treatment plan lived in the front desk’s head, and the chair stood idle.

Three practices, twelve chairs, 30–40 staff. The dentist drew up a treatment plan at the appointment and from then on it survived on memory: the front desk kept a log and called patients whenever there was time. Around half of patients never reached the end of a plan — not because they changed their minds, but because nobody reminded them of the next stage in time. Weeks passed between stages where a few days would have done clinically. Chairs were allocated by dentist rather than by procedure, so a long piece of work could not find a window while short visits chopped up the day. Lab work was ordered by phone and tracked in a spreadsheet: whoever placed the order knew when the crown would arrive, and appointments were booked with a safety margin. The price of a plan, the split of payments across stages and any instalment arrangement were all worked out by hand.

Planning from the chair, not from the logbook.

Same stack as our other projects: Python and Django at the core, PostgreSQL as the source of truth for the patient, the stage and the chair. The platform knows how long each procedure takes, what equipment it needs and how long the external dental lab takes, and builds the schedule from that rather than from one dentist’s free time. The lab is connected by an order and status exchange, so turnaround time is an input to planning exactly like procedure length.

Six loops in two months.

The order was chosen by money: first what brings a patient back and fills a chair, then everything else.

Before and after.

We planned from the chair, not from the dentist.

Three practices, twelve chairs. A treatment plan stretches over months, but a chair stands idle right now, and it is the chair that caps revenue rather than the number of dentists. So the schedule is assembled from chairs and equipment: the system knows how long a procedure takes, what it needs and when the lab work arrives, and fits visits together so a gap is closed by the next suitable stage rather than by whichever patient happens to be free. The 15–20% came without hiring anyone and without adding a chair. The whole loop took about two months: a dental practice is process-wise simpler than a retailer, and the lab was the only outside system we had to agree an exchange with.