Why We Built Virtual Visits as a Second Channel, Not the First

It would have been faster to build video calling first. A built-in video call is a well-understood engineering problem, and it sidesteps GPS drift, multi-storey buildings, and every other headache of physical verification.

We did it in the other order anyway, and the reason comes down to what compliance teams actually trust.

Trust has to be earned in person first

A verified in-person visit, backed by GPS and a time-boxed QR scan at a real clinic, is about as close to unambiguous proof as field verification gets. Building that first meant our earliest pilots were judged against the hardest standard we could set for ourselves, rather than the easiest. If a compliance head was going to trust a new verification system at all, we wanted their first experience of it to be the strongest possible evidence, not the most convenient one.

Virtual visits solve a real problem, but it is a narrower one. Specialists in tier-1 cities with packed OPD schedules, doctors in regions where travel time genuinely does not justify a 15-minute in-person slot, and rescheduled visits that would otherwise be lost entirely, these are good reasons to have a video channel. They are not a good reason to make video the default.

What virtual visits actually require

• •A built-in video call, so the interaction happens inside the platform rather than over a personal number or a third-party app with no record
• •Explicit consent captured from both the doctor and the rep before the call is treated as a valid visit, not implied by simply joining
• •The same AI note extraction and compliance flagging applied to in-person visits, so a virtual visit produces an equally structured record
• •Clear labeling in every downstream report of which visits were in-person and which were virtual, because compliance teams reasonably want to know the difference

Virtual visits are live in the product today as a genuine second channel, not a watered-down fallback. But we are glad we made compliance teams trust the harder version of this problem first. It made every subsequent conversation about the video channel easier, because the foundation underneath it had already proven itself.