Tells health workers which TB and BP patients to call about missed doses
TL;DR Doseminder pulls together cheap signs that a patient has stopped taking their medicines (missed-call check-ins, late pharmacy refills, pillbox openings) and hands health workers a short morning list of who to call first. TB programmes, hospitals and pharma support schemes pay per patient. It swaps out the seed's sensor pills, which flopped as a business.
India diagnosed 26.18 lakh TB patients in 2024[1]. In one cohort of 2,006 patients, 18% ended badly, including 6.8% lost to follow-up[2].
Chronic disease looks worse. A 2026 review of 16 Indian studies found 39.3% of people with type 2 diabetes weren't taking medicines as prescribed[3], and 27% of patients in the national hypertension programme had no recorded visit in the reporting period[4].
So who has stopped? That's the hard bit. In a study of 650 TB patients, no common measure (missed calls, pill counts or simply asking) caught more than 70% of those skipping doses when checked against a urine drug test[5]. Health workers end up chasing everyone, or nobody.
What it does, in regulatory terms: adherence monitoring only. Doseminder does not diagnose, treat or measure drug levels, and clinical decisions stay with the treating doctor. Software that prompts clinicians to act on a serious condition like TB is likely Class A or B under CDSCO's 2026 guidance[6], so get a written view early.
Each patient gets whichever signals suit their phone and programme - a missed-call or voice check-in in their language, pharmacy refill records, a smart pillbox for the highest-risk cases and the odd urine test at TB home visits. A model scores each patient's chance of dropping off this week and says why (three missed check-ins, a refill two weeks late). Workers get a ranked list with a call script, and for TB the data flows into the government's Ni-kshay platform.
The seed put a sensor in every pill. The main maker of sensor pills went bankrupt in 2020 after raising about US$500M[7], and every sensor tablet would need its own drug approval here. Cheap signals put together get you to the same place.
Patients are adults on long courses, six months or more for TB and life for hypertension and diabetes. Many are daily-wage workers with basic phones. Health workers use the list but someone else pays - state TB cells and district officers (slow, often donor-funded at first), private hospitals with chronic-care clinics (faster) and pharma patient-support schemes. My rough guess is ₹30–100 per patient per month. A smart pillbox cost about ₹10,000 in 2017[8], so boxes go only to high-risk patients.
India already runs digital adherence at scale. The TB programme linked a missed-call dosing system to Ni-kshay in 2018, and it has reached over 5 lakh people[9]. Drug-resistant TB teams use a smart pillbox. The gap is just as clear though - in Himachal Pradesh the missed-call system didn't change outcomes, partly because only 33% of patients ever called[9]. CDSCO's July 2026 software guidance now sets the rules for monitoring software[6].
There's no Indian market for sensor-pill adherence and no published figure for adherence services. So count patients instead. India logs 26.18 lakh TB diagnoses a year[1] and has more than 20 lakh people enrolled in the hypertension programme[4]. At ₹30–100 per patient per month, reaching 10% of those two groups would bring in roughly ₹17–55 Cr a year by my arithmetic. Estimates of India's home monitoring devices differ about 17-fold, so I wouldn't lean on them.
The TB programme's missed-call system hides a toll-free number behind each dose and flags patients who stop calling[9]. It's cheap and widespread, but it's one signal. Smart pillboxes log openings for drug-resistant TB, and patients have complained about flimsy builds and loud alarms[10]. Sensor-pill makers and their pharma partners proved the pills work in trials but not as a business[7].
Nobody combines signals and ranks risk for Indian programmes. That's the gap.
| Difficulty | Hard (52/100) |
| Starting capital | ₹1.9–3.45 Cr |
| Time to launch | 6–9 months |
The software is manageable. Selling is the hard part, because public programmes buy slowly, usually after a donor-funded pilot, and hospitals want proof before they pay per patient. You also need historical patient data to train a risk model, which means data agreements and a fair bit of patience.
Possibly not, but get it in writing. A plain reminder app may fall outside device rules, and clinical software for TB probably doesn't.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't let AI decide treatment or label a patient as non-compliant without a human check. Signals can be wrong, and the treating doctor makes every clinical call.
Want to work out the AI layer for Doseminder? Talk to Yash →
Here's what can go south -
A combined-signal adherence service meets a real need. A sensor-pill company is not a startup business in India, full stop. If you know public health programmes or hospital operations and can sit through slow procurement then this fits you. If you're a tech-first founder then read how the big sensor-pill company went under before anything else. Prove first that risk-ranked calls cut missed visits in one clinic.
My estimate is ₹1.9–3.45 Cr to launch, covering the software, voice check-ins, smart pillboxes for high-risk patients and early pilots. The kit breaks it down line by line.
Maybe. A plain reminder app may sit outside device rules, but software that prompts clinicians on TB is likely Class A or B, so get a written CDSCO view.
Yes, missed doses are a problem everywhere. Check local device software rules and health data laws first.
The free Doseminder validation kit has the line-by-line capital breakdown, the effort scoring, the software-classification and data-protection checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for programme staff and hospitals, and 5 assumptions to test. Download the kit →
Thinking about building Doseminder or something close to it? The make-or-break work is landing a first paying hospital and a donor-backed district pilot that measures an outcome buyers care about, and that's the kind of thing I help founders with. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.