Healthcare · 2025
How a Noida hospital cut patient wait times by 62%
40-bed multi-speciality hospital
Illustrative sample — not a real client engagement
These projects demonstrate the kind of work and the level of detail a real case study will carry. They are composites, not client work, and carry no client attribution. Real engagements replace them as clients give written permission.
- 62%
- Reduction in average wait
- 94%
- Reduction in double-booking
- 3.5 hrs
- Reception time saved daily
- 100%
- Clinician adoption at 60 days
74 min → 28 min, measured over 8 weeks pre and post go-live
Slot conflicts recorded by the system vs. register audit
Time-and-motion sample across two reception staff
Consultations closed in-system vs. total consultations
At a glance
- Industry
- Healthcare
- Timeline
- 5 months
- Team size
- 5
- Investment
- ₹34,00,000
- Payback
- 8 months
The problem
Scheduling ran on paper registers across four departments. Around 40% of slots were double-booked because no department could see another's diary, and the average patient waited 74 minutes past their appointment time. Reception absorbed the complaints; nobody had data to fix the cause.
The challenge
The hospital could not stop operating during a rollout, and clinicians had already abandoned one previous software attempt because it added clicks during consultation. Any system also had to work when the internet dropped, which happened several times a week.
Our approach
We spent the first two weeks shadowing reception and three consulting rooms rather than gathering requirements in a meeting. That surfaced the real constraint: doctors would tolerate almost nothing during a consultation. So we designed the clinical screen first, capped at three interactions to close a visit, and built scheduling around it — rather than the reverse, which is how the previous attempt had failed.
What we built
The solution.
Unified scheduling
One slot ledger across four departments, with conflict prevention at booking rather than detection afterwards.
Offline-first reception
Registration and billing run against local storage and sync on reconnect, so a connectivity drop never stops the queue.
Three-tap consultation
Prescription templates and favourites per doctor, so closing a visit takes seconds rather than a form.
Live queue display
Waiting-area screens showing real position, which reduced reception interruptions substantially on its own.
Technology used
- Next.js
- Node.js
- PostgreSQL
- Redis
- AWS
Timeline
How it ran.
Discovery
Weeks 1–2Shadowing reception and consulting rooms; measuring the existing baseline.Design & prototype
Weeks 3–5Clinical screen tested with four doctors before any engineering.Build
Weeks 6–16Fortnightly demos with reception staff in the room.Parallel run
Weeks 17–19Paper and system operating together until the numbers agreed.Go-live
Week 20Department by department over five days, with rollback available.Return
Did it pay for itself?
Investment against measured annual return. Figures agreed with the client's finance team rather than estimated by us.
- ₹34,00,000
- Investment
- ₹51,00,000
- Annual return
- 8 months
- Payback period
What we'd do differently
The part that didn't go to plan.
We under-scoped data migration. Fifteen years of paper records were far messier than the sample we assessed, and migration ran three weeks over. On a similar project now we'd audit a random sample of the oldest records during discovery rather than the most recent, because the recent ones are always the tidiest.
What's next
Pharmacy and lab modules are in progress, and ABDM linkage is scheduled for the next phase.
Get results like these.
Thirty minutes with an engineer. We'll tell you honestly whether your problem is the same shape as this one.