The ledger that learns your handwriting
Photograph your visit slip and AI turns it into a record in seconds. Correct anything it misread, and it remembers — so your clinic's shorthand gets read better every week. Then chase what you're owed on WhatsApp, with a UPI link for the exact amount.
Generic software has never read a dentist's slip
"The cap colour was wrong" — and the only record is a slip in a drawer, if it's still there.
Patients leave part-paid and settle slowly. Chasing means remembering who owes what, and for how long.
Every other tool hands you blank fields. You are the data-entry operator, at 9pm, from memory.
Every correction makes the next slip easier
Most AI tools give every clinic the same generic model. ClinicKhata keeps a profile for yourclinic: your abbreviations, your fee patterns, the way you write a tooth number. You never train anything — you just fix what's wrong before saving, and that fix becomes the rule.
The same slip you already write at the chair.
Patient, tooth, treatment, medicine, cost, payment.
Two taps before saving — the review you'd do anyway.
Automatically. Nothing for you to manage or configure.
Snap · Check · Recover
Between patients, one photo. Several photos for one visit if the notes run long.
Nothing saves until you've seen it. Your edits teach your clinic's profile as a side effect.
Aged dues on the home screen; one tap sends a WhatsApp reminder with a UPI link.
A record that exists to get you paid
The point of the ledger is the money in it. Home opens on total outstanding, aged 0–30 / 31–60 / 60+ days, and what you've recovered this month. Every patient row carries a WhatsApp reminder with a UPI deep-link set to the exact balance — in Hindi or English.
Stored in India. Never training data.
Your patients' records are stored in India, fenced off per clinic at the database and encrypted at rest, with audit logs and erasure on request — aligned to India's DPDPA. And the learning loop only sharpens your clinic's own prompt: your corrections are never pooled with other clinics, never sold, and never used to train an AI model.
Tuned against one real clinic's slips
ClinicKhata is built and tested alongside a working dental clinic in India — the same handwriting, the same shorthand, the same end-of-day tiredness. The learning loop exists because we watched how often a generic model got her slips slightly wrong.
We're early and we'd rather say so plainly: no customer logos, no invented testimonials.
Create your clinic
Set up in a couple of minutes: enter your clinic details, verify your email with a one-time code, and you land straight on your dashboard. You can start photographing slips right away.
Questions dentists ask
No. You review each record before saving anyway — the corrections you make there are the training. There's no separate setup step.
No. Slip photos, patient records and your corrections are never used as training data for any AI model. Corrections adjust the extraction prompt for your clinic alone — never pooled with other clinics, never shared, never sold. Your records are stored in India, aligned to the DPDPA.
Yes, both. Reminders send from your own WhatsApp; the link opens the patient's UPI app. They install nothing.
Not yet. Today it's the slip photo, with manual typing as a fallback. Dictation may come later.
No. ClinicKhata records visits that already happened. Scheduling is a different product.
No. UPI, WhatsApp and Hindi are built into how it works, and your records are stored in India — it's for Indian clinics by design.